Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
/*-------------------------------------------------------------------------
|
|
|
|
|
*
|
2024-08-05 09:35:42 -04:00
|
|
|
* astreamer_file.c
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
*
|
2024-08-07 08:49:41 -04:00
|
|
|
* Archive streamers that write to files. astreamer_plain_writer writes
|
|
|
|
|
* the whole archive to a single file, and astreamer_extractor writes
|
|
|
|
|
* each archive member to a separate file in a given directory.
|
|
|
|
|
*
|
2025-01-01 11:21:55 -05:00
|
|
|
* Portions Copyright (c) 1996-2025, PostgreSQL Global Development Group
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
*
|
|
|
|
|
* IDENTIFICATION
|
2024-12-23 01:41:49 -05:00
|
|
|
* src/fe_utils/astreamer_file.c
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
*-------------------------------------------------------------------------
|
|
|
|
|
*/
|
|
|
|
|
|
|
|
|
|
#include "postgres_fe.h"
|
|
|
|
|
|
|
|
|
|
#include <unistd.h>
|
|
|
|
|
|
|
|
|
|
#include "common/file_perm.h"
|
2024-03-13 10:07:00 -04:00
|
|
|
#include "common/logging.h"
|
2024-08-05 11:40:29 -04:00
|
|
|
#include "fe_utils/astreamer.h"
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
|
2024-08-05 09:35:42 -04:00
|
|
|
typedef struct astreamer_plain_writer
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
{
|
2024-08-05 09:35:42 -04:00
|
|
|
astreamer base;
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
char *pathname;
|
|
|
|
|
FILE *file;
|
|
|
|
|
bool should_close_file;
|
2024-08-05 09:35:42 -04:00
|
|
|
} astreamer_plain_writer;
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
|
2024-08-05 09:35:42 -04:00
|
|
|
typedef struct astreamer_extractor
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
{
|
2024-08-05 09:35:42 -04:00
|
|
|
astreamer base;
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
char *basepath;
|
|
|
|
|
const char *(*link_map) (const char *);
|
|
|
|
|
void (*report_output_file) (const char *);
|
|
|
|
|
char filename[MAXPGPATH];
|
|
|
|
|
FILE *file;
|
2024-08-05 09:35:42 -04:00
|
|
|
} astreamer_extractor;
|
|
|
|
|
|
|
|
|
|
static void astreamer_plain_writer_content(astreamer *streamer,
|
|
|
|
|
astreamer_member *member,
|
|
|
|
|
const char *data, int len,
|
|
|
|
|
astreamer_archive_context context);
|
|
|
|
|
static void astreamer_plain_writer_finalize(astreamer *streamer);
|
|
|
|
|
static void astreamer_plain_writer_free(astreamer *streamer);
|
|
|
|
|
|
|
|
|
|
static const astreamer_ops astreamer_plain_writer_ops = {
|
|
|
|
|
.content = astreamer_plain_writer_content,
|
|
|
|
|
.finalize = astreamer_plain_writer_finalize,
|
|
|
|
|
.free = astreamer_plain_writer_free
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
};
|
|
|
|
|
|
2024-08-05 09:35:42 -04:00
|
|
|
static void astreamer_extractor_content(astreamer *streamer,
|
|
|
|
|
astreamer_member *member,
|
|
|
|
|
const char *data, int len,
|
|
|
|
|
astreamer_archive_context context);
|
|
|
|
|
static void astreamer_extractor_finalize(astreamer *streamer);
|
|
|
|
|
static void astreamer_extractor_free(astreamer *streamer);
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
static void extract_directory(const char *filename, mode_t mode);
|
|
|
|
|
static void extract_link(const char *filename, const char *linktarget);
|
|
|
|
|
static FILE *create_file_for_extract(const char *filename, mode_t mode);
|
|
|
|
|
|
2024-08-05 09:35:42 -04:00
|
|
|
static const astreamer_ops astreamer_extractor_ops = {
|
|
|
|
|
.content = astreamer_extractor_content,
|
|
|
|
|
.finalize = astreamer_extractor_finalize,
|
|
|
|
|
.free = astreamer_extractor_free
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
};
|
|
|
|
|
|
|
|
|
|
/*
|
2024-08-05 09:35:42 -04:00
|
|
|
* Create a astreamer that just writes data to a file.
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
*
|
|
|
|
|
* The caller must specify a pathname and may specify a file. The pathname is
|
|
|
|
|
* used for error-reporting purposes either way. If file is NULL, the pathname
|
|
|
|
|
* also identifies the file to which the data should be written: it is opened
|
|
|
|
|
* for writing and closed when done. If file is not NULL, the data is written
|
|
|
|
|
* there.
|
|
|
|
|
*/
|
2024-08-05 09:35:42 -04:00
|
|
|
astreamer *
|
|
|
|
|
astreamer_plain_writer_new(char *pathname, FILE *file)
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
{
|
2024-08-05 09:35:42 -04:00
|
|
|
astreamer_plain_writer *streamer;
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
|
2024-08-05 09:35:42 -04:00
|
|
|
streamer = palloc0(sizeof(astreamer_plain_writer));
|
|
|
|
|
*((const astreamer_ops **) &streamer->base.bbs_ops) =
|
|
|
|
|
&astreamer_plain_writer_ops;
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
|
|
|
|
|
streamer->pathname = pstrdup(pathname);
|
|
|
|
|
streamer->file = file;
|
|
|
|
|
|
|
|
|
|
if (file == NULL)
|
|
|
|
|
{
|
|
|
|
|
streamer->file = fopen(pathname, "wb");
|
|
|
|
|
if (streamer->file == NULL)
|
2022-04-08 14:55:14 -04:00
|
|
|
pg_fatal("could not create file \"%s\": %m", pathname);
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
streamer->should_close_file = true;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
return &streamer->base;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
* Write archive content to file.
|
|
|
|
|
*/
|
|
|
|
|
static void
|
2024-08-05 09:35:42 -04:00
|
|
|
astreamer_plain_writer_content(astreamer *streamer,
|
|
|
|
|
astreamer_member *member, const char *data,
|
|
|
|
|
int len, astreamer_archive_context context)
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
{
|
2024-08-05 09:35:42 -04:00
|
|
|
astreamer_plain_writer *mystreamer;
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
|
2024-08-05 09:35:42 -04:00
|
|
|
mystreamer = (astreamer_plain_writer *) streamer;
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
|
|
|
|
|
if (len == 0)
|
|
|
|
|
return;
|
|
|
|
|
|
|
|
|
|
errno = 0;
|
|
|
|
|
if (fwrite(data, len, 1, mystreamer->file) != 1)
|
|
|
|
|
{
|
|
|
|
|
/* if write didn't set errno, assume problem is no disk space */
|
|
|
|
|
if (errno == 0)
|
|
|
|
|
errno = ENOSPC;
|
2022-04-08 14:55:14 -04:00
|
|
|
pg_fatal("could not write to file \"%s\": %m",
|
|
|
|
|
mystreamer->pathname);
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
* End-of-archive processing when writing to a plain file consists of closing
|
|
|
|
|
* the file if we opened it, but not if the caller provided it.
|
|
|
|
|
*/
|
|
|
|
|
static void
|
2024-08-05 09:35:42 -04:00
|
|
|
astreamer_plain_writer_finalize(astreamer *streamer)
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
{
|
2024-08-05 09:35:42 -04:00
|
|
|
astreamer_plain_writer *mystreamer;
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
|
2024-08-05 09:35:42 -04:00
|
|
|
mystreamer = (astreamer_plain_writer *) streamer;
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
|
|
|
|
|
if (mystreamer->should_close_file && fclose(mystreamer->file) != 0)
|
2022-04-08 14:55:14 -04:00
|
|
|
pg_fatal("could not close file \"%s\": %m",
|
|
|
|
|
mystreamer->pathname);
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
|
|
|
|
|
mystreamer->file = NULL;
|
|
|
|
|
mystreamer->should_close_file = false;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/*
|
2024-08-05 09:35:42 -04:00
|
|
|
* Free memory associated with this astreamer.
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
*/
|
|
|
|
|
static void
|
2024-08-05 09:35:42 -04:00
|
|
|
astreamer_plain_writer_free(astreamer *streamer)
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
{
|
2024-08-05 09:35:42 -04:00
|
|
|
astreamer_plain_writer *mystreamer;
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
|
2024-08-05 09:35:42 -04:00
|
|
|
mystreamer = (astreamer_plain_writer *) streamer;
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
|
|
|
|
|
Assert(!mystreamer->should_close_file);
|
|
|
|
|
Assert(mystreamer->base.bbs_next == NULL);
|
|
|
|
|
|
|
|
|
|
pfree(mystreamer->pathname);
|
|
|
|
|
pfree(mystreamer);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/*
|
2024-08-05 09:35:42 -04:00
|
|
|
* Create a astreamer that extracts an archive.
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
*
|
|
|
|
|
* All pathnames in the archive are interpreted relative to basepath.
|
|
|
|
|
*
|
2024-08-05 09:35:42 -04:00
|
|
|
* Unlike e.g. astreamer_plain_writer_new() we can't do anything useful here
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
* with untyped chunks; we need typed chunks which follow the rules described
|
2024-08-05 09:35:42 -04:00
|
|
|
* in astreamer.h. Assuming we have that, we don't need to worry about the
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
* original archive format; it's enough to just look at the member information
|
|
|
|
|
* provided and write to the corresponding file.
|
|
|
|
|
*
|
|
|
|
|
* 'link_map' is a function that will be applied to the target of any
|
|
|
|
|
* symbolic link, and which should return a replacement pathname to be used
|
|
|
|
|
* in its place. If NULL, the symbolic link target is used without
|
|
|
|
|
* modification.
|
|
|
|
|
*
|
|
|
|
|
* 'report_output_file' is a function that will be called each time we open a
|
|
|
|
|
* new output file. The pathname to that file is passed as an argument. If
|
|
|
|
|
* NULL, the call is skipped.
|
|
|
|
|
*/
|
2024-08-05 09:35:42 -04:00
|
|
|
astreamer *
|
|
|
|
|
astreamer_extractor_new(const char *basepath,
|
|
|
|
|
const char *(*link_map) (const char *),
|
|
|
|
|
void (*report_output_file) (const char *))
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
{
|
2024-08-05 09:35:42 -04:00
|
|
|
astreamer_extractor *streamer;
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
|
2024-08-05 09:35:42 -04:00
|
|
|
streamer = palloc0(sizeof(astreamer_extractor));
|
|
|
|
|
*((const astreamer_ops **) &streamer->base.bbs_ops) =
|
|
|
|
|
&astreamer_extractor_ops;
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
streamer->basepath = pstrdup(basepath);
|
|
|
|
|
streamer->link_map = link_map;
|
|
|
|
|
streamer->report_output_file = report_output_file;
|
|
|
|
|
|
|
|
|
|
return &streamer->base;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
* Extract archive contents to the filesystem.
|
|
|
|
|
*/
|
|
|
|
|
static void
|
2024-08-05 09:35:42 -04:00
|
|
|
astreamer_extractor_content(astreamer *streamer, astreamer_member *member,
|
|
|
|
|
const char *data, int len,
|
|
|
|
|
astreamer_archive_context context)
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
{
|
2024-08-05 09:35:42 -04:00
|
|
|
astreamer_extractor *mystreamer = (astreamer_extractor *) streamer;
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
int fnamelen;
|
|
|
|
|
|
2024-08-05 09:35:42 -04:00
|
|
|
Assert(member != NULL || context == ASTREAMER_ARCHIVE_TRAILER);
|
|
|
|
|
Assert(context != ASTREAMER_UNKNOWN);
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
|
|
|
|
|
switch (context)
|
|
|
|
|
{
|
2024-08-05 09:35:42 -04:00
|
|
|
case ASTREAMER_MEMBER_HEADER:
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
Assert(mystreamer->file == NULL);
|
|
|
|
|
|
|
|
|
|
/* Prepend basepath. */
|
|
|
|
|
snprintf(mystreamer->filename, sizeof(mystreamer->filename),
|
|
|
|
|
"%s/%s", mystreamer->basepath, member->pathname);
|
|
|
|
|
|
|
|
|
|
/* Remove any trailing slash. */
|
|
|
|
|
fnamelen = strlen(mystreamer->filename);
|
|
|
|
|
if (mystreamer->filename[fnamelen - 1] == '/')
|
|
|
|
|
mystreamer->filename[fnamelen - 1] = '\0';
|
|
|
|
|
|
|
|
|
|
/* Dispatch based on file type. */
|
|
|
|
|
if (member->is_directory)
|
|
|
|
|
extract_directory(mystreamer->filename, member->mode);
|
|
|
|
|
else if (member->is_link)
|
|
|
|
|
{
|
|
|
|
|
const char *linktarget = member->linktarget;
|
|
|
|
|
|
|
|
|
|
if (mystreamer->link_map)
|
|
|
|
|
linktarget = mystreamer->link_map(linktarget);
|
|
|
|
|
extract_link(mystreamer->filename, linktarget);
|
|
|
|
|
}
|
|
|
|
|
else
|
|
|
|
|
mystreamer->file =
|
|
|
|
|
create_file_for_extract(mystreamer->filename,
|
|
|
|
|
member->mode);
|
|
|
|
|
|
|
|
|
|
/* Report output file change. */
|
|
|
|
|
if (mystreamer->report_output_file)
|
|
|
|
|
mystreamer->report_output_file(mystreamer->filename);
|
|
|
|
|
break;
|
|
|
|
|
|
2024-08-05 09:35:42 -04:00
|
|
|
case ASTREAMER_MEMBER_CONTENTS:
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
if (mystreamer->file == NULL)
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
errno = 0;
|
|
|
|
|
if (len > 0 && fwrite(data, len, 1, mystreamer->file) != 1)
|
|
|
|
|
{
|
|
|
|
|
/* if write didn't set errno, assume problem is no disk space */
|
|
|
|
|
if (errno == 0)
|
|
|
|
|
errno = ENOSPC;
|
2022-04-08 14:55:14 -04:00
|
|
|
pg_fatal("could not write to file \"%s\": %m",
|
|
|
|
|
mystreamer->filename);
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
}
|
|
|
|
|
break;
|
|
|
|
|
|
2024-08-05 09:35:42 -04:00
|
|
|
case ASTREAMER_MEMBER_TRAILER:
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
if (mystreamer->file == NULL)
|
|
|
|
|
break;
|
|
|
|
|
fclose(mystreamer->file);
|
|
|
|
|
mystreamer->file = NULL;
|
|
|
|
|
break;
|
|
|
|
|
|
2024-08-05 09:35:42 -04:00
|
|
|
case ASTREAMER_ARCHIVE_TRAILER:
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
default:
|
|
|
|
|
/* Shouldn't happen. */
|
2022-04-08 14:55:14 -04:00
|
|
|
pg_fatal("unexpected state while extracting archive");
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
Fix pg_basebackup with in-place tablespaces some more.
Commit c6f2f01611d4f2c412e92eb7893f76fa590818e8 purported to make
this work, but problems remained. In a plain-format backup, the
files from an in-place tablespace got included in the tar file for
the main tablespace, which is wrong but it's not clear that it
has any user-visible consequences. In a tar-format backup, the
TABLESPACE_MAP option is used, and so we never iterated over
pg_tblspc and thus never backed up the in-place tablespaces
anywhere at all.
To fix this, reverse the changes in that commit, so that when we scan
pg_tblspc during a backup, we create tablespaceinfo objects even for
in-place tablespaces. We set the field that would normally contain the
absolute pathname to the relative path pg_tblspc/${TSOID}, and that's
good enough to make basebackup.c happy without any further changes.
However, pg_basebackup needs a couple of adjustments to make it work.
First, it needs to understand that a relative path for a tablespace
means it's an in-place tablespace. Second, it needs to tolerate the
situation where restoring the main tablespace tries to create
pg_tblspc or a subdirectory and finds that it already exists, because
we restore user-defined tablespaces before the main tablespace.
Since in-place tablespaces are only intended for use in development
and testing, no back-patch.
Patch by me, reviewed by Thomas Munro and Michael Paquier.
Discussion: http://postgr.es/m/CA+TgmobwvbEp+fLq2PykMYzizcvuNv0a7gPMJtxOTMOuuRLMHg@mail.gmail.com
2023-04-18 11:23:34 -04:00
|
|
|
/*
|
|
|
|
|
* Should we tolerate an already-existing directory?
|
|
|
|
|
*
|
|
|
|
|
* When streaming WAL, pg_wal (or pg_xlog for pre-9.6 clusters) will have been
|
|
|
|
|
* created by the wal receiver process. Also, when the WAL directory location
|
|
|
|
|
* was specified, pg_wal (or pg_xlog) has already been created as a symbolic
|
|
|
|
|
* link before starting the actual backup. So just ignore creation failures
|
|
|
|
|
* on related directories.
|
|
|
|
|
*
|
|
|
|
|
* If in-place tablespaces are used, pg_tblspc and subdirectories may already
|
|
|
|
|
* exist when we get here. So tolerate that case, too.
|
|
|
|
|
*/
|
|
|
|
|
static bool
|
|
|
|
|
should_allow_existing_directory(const char *pathname)
|
|
|
|
|
{
|
|
|
|
|
const char *filename = last_dir_separator(pathname) + 1;
|
|
|
|
|
|
|
|
|
|
if (strcmp(filename, "pg_wal") == 0 ||
|
|
|
|
|
strcmp(filename, "pg_xlog") == 0 ||
|
|
|
|
|
strcmp(filename, "archive_status") == 0 ||
|
2023-12-20 09:49:12 -05:00
|
|
|
strcmp(filename, "summaries") == 0 ||
|
Fix pg_basebackup with in-place tablespaces some more.
Commit c6f2f01611d4f2c412e92eb7893f76fa590818e8 purported to make
this work, but problems remained. In a plain-format backup, the
files from an in-place tablespace got included in the tar file for
the main tablespace, which is wrong but it's not clear that it
has any user-visible consequences. In a tar-format backup, the
TABLESPACE_MAP option is used, and so we never iterated over
pg_tblspc and thus never backed up the in-place tablespaces
anywhere at all.
To fix this, reverse the changes in that commit, so that when we scan
pg_tblspc during a backup, we create tablespaceinfo objects even for
in-place tablespaces. We set the field that would normally contain the
absolute pathname to the relative path pg_tblspc/${TSOID}, and that's
good enough to make basebackup.c happy without any further changes.
However, pg_basebackup needs a couple of adjustments to make it work.
First, it needs to understand that a relative path for a tablespace
means it's an in-place tablespace. Second, it needs to tolerate the
situation where restoring the main tablespace tries to create
pg_tblspc or a subdirectory and finds that it already exists, because
we restore user-defined tablespaces before the main tablespace.
Since in-place tablespaces are only intended for use in development
and testing, no back-patch.
Patch by me, reviewed by Thomas Munro and Michael Paquier.
Discussion: http://postgr.es/m/CA+TgmobwvbEp+fLq2PykMYzizcvuNv0a7gPMJtxOTMOuuRLMHg@mail.gmail.com
2023-04-18 11:23:34 -04:00
|
|
|
strcmp(filename, "pg_tblspc") == 0)
|
|
|
|
|
return true;
|
|
|
|
|
|
|
|
|
|
if (strspn(filename, "0123456789") == strlen(filename))
|
|
|
|
|
{
|
|
|
|
|
const char *pg_tblspc = strstr(pathname, "/pg_tblspc/");
|
|
|
|
|
|
|
|
|
|
return pg_tblspc != NULL && pg_tblspc + 11 == filename;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
return false;
|
|
|
|
|
}
|
|
|
|
|
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
/*
|
|
|
|
|
* Create a directory.
|
|
|
|
|
*/
|
|
|
|
|
static void
|
|
|
|
|
extract_directory(const char *filename, mode_t mode)
|
|
|
|
|
{
|
Fix pg_basebackup with in-place tablespaces some more.
Commit c6f2f01611d4f2c412e92eb7893f76fa590818e8 purported to make
this work, but problems remained. In a plain-format backup, the
files from an in-place tablespace got included in the tar file for
the main tablespace, which is wrong but it's not clear that it
has any user-visible consequences. In a tar-format backup, the
TABLESPACE_MAP option is used, and so we never iterated over
pg_tblspc and thus never backed up the in-place tablespaces
anywhere at all.
To fix this, reverse the changes in that commit, so that when we scan
pg_tblspc during a backup, we create tablespaceinfo objects even for
in-place tablespaces. We set the field that would normally contain the
absolute pathname to the relative path pg_tblspc/${TSOID}, and that's
good enough to make basebackup.c happy without any further changes.
However, pg_basebackup needs a couple of adjustments to make it work.
First, it needs to understand that a relative path for a tablespace
means it's an in-place tablespace. Second, it needs to tolerate the
situation where restoring the main tablespace tries to create
pg_tblspc or a subdirectory and finds that it already exists, because
we restore user-defined tablespaces before the main tablespace.
Since in-place tablespaces are only intended for use in development
and testing, no back-patch.
Patch by me, reviewed by Thomas Munro and Michael Paquier.
Discussion: http://postgr.es/m/CA+TgmobwvbEp+fLq2PykMYzizcvuNv0a7gPMJtxOTMOuuRLMHg@mail.gmail.com
2023-04-18 11:23:34 -04:00
|
|
|
if (mkdir(filename, pg_dir_create_mode) != 0 &&
|
|
|
|
|
(errno != EEXIST || !should_allow_existing_directory(filename)))
|
|
|
|
|
pg_fatal("could not create directory \"%s\": %m",
|
|
|
|
|
filename);
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
|
|
|
|
|
#ifndef WIN32
|
|
|
|
|
if (chmod(filename, mode))
|
2022-04-08 14:55:14 -04:00
|
|
|
pg_fatal("could not set permissions on directory \"%s\": %m",
|
|
|
|
|
filename);
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
#endif
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
* Create a symbolic link.
|
|
|
|
|
*
|
|
|
|
|
* It's most likely a link in pg_tblspc directory, to the location of a
|
|
|
|
|
* tablespace. Apply any tablespace mapping given on the command line
|
|
|
|
|
* (--tablespace-mapping). (We blindly apply the mapping without checking that
|
|
|
|
|
* the link really is inside pg_tblspc. We don't expect there to be other
|
|
|
|
|
* symlinks in a data directory, but if there are, you can call it an
|
|
|
|
|
* undocumented feature that you can map them too.)
|
|
|
|
|
*/
|
|
|
|
|
static void
|
|
|
|
|
extract_link(const char *filename, const char *linktarget)
|
|
|
|
|
{
|
|
|
|
|
if (symlink(linktarget, filename) != 0)
|
2022-04-08 14:55:14 -04:00
|
|
|
pg_fatal("could not create symbolic link from \"%s\" to \"%s\": %m",
|
|
|
|
|
filename, linktarget);
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
* Create a regular file.
|
|
|
|
|
*
|
|
|
|
|
* Return the resulting handle so we can write the content to the file.
|
|
|
|
|
*/
|
|
|
|
|
static FILE *
|
|
|
|
|
create_file_for_extract(const char *filename, mode_t mode)
|
|
|
|
|
{
|
|
|
|
|
FILE *file;
|
|
|
|
|
|
|
|
|
|
file = fopen(filename, "wb");
|
|
|
|
|
if (file == NULL)
|
2022-04-08 14:55:14 -04:00
|
|
|
pg_fatal("could not create file \"%s\": %m", filename);
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
|
|
|
|
|
#ifndef WIN32
|
|
|
|
|
if (chmod(filename, mode))
|
2022-04-08 14:55:14 -04:00
|
|
|
pg_fatal("could not set permissions on file \"%s\": %m",
|
|
|
|
|
filename);
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
#endif
|
|
|
|
|
|
|
|
|
|
return file;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
* End-of-stream processing for extracting an archive.
|
|
|
|
|
*
|
|
|
|
|
* There's nothing to do here but sanity checking.
|
|
|
|
|
*/
|
|
|
|
|
static void
|
2024-08-05 09:35:42 -04:00
|
|
|
astreamer_extractor_finalize(astreamer *streamer)
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
{
|
2024-08-05 09:35:42 -04:00
|
|
|
astreamer_extractor *mystreamer PG_USED_FOR_ASSERTS_ONLY
|
|
|
|
|
= (astreamer_extractor *) streamer;
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
|
|
|
|
|
Assert(mystreamer->file == NULL);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
* Free memory.
|
|
|
|
|
*/
|
|
|
|
|
static void
|
2024-08-05 09:35:42 -04:00
|
|
|
astreamer_extractor_free(astreamer *streamer)
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
{
|
2024-08-05 09:35:42 -04:00
|
|
|
astreamer_extractor *mystreamer = (astreamer_extractor *) streamer;
|
Introduce 'bbstreamer' abstraction to modularize pg_basebackup.
pg_basebackup knows how to do quite a few things with a backup that it
gets from the server, like just write out the files, or compress them
first, or even parse the tar format and inject a modified
postgresql.auto.conf file into the archive generated by the server.
Unforatunely, this makes pg_basebackup.c a very large source file, and
also somewhat difficult to enhance, because for example the knowledge
that the server is sending us a 'tar' file rather than some other sort
of archive is spread all over the place rather than centralized.
In an effort to improve this situation, this commit invents a new
'bbstreamer' abstraction. Each archive received from the server is
fed to a bbstreamer which may choose to dispose of it or pass it
along to some other bbstreamer. Chunks may also be "labelled"
according to whether they are part of the payload data of a file
in the archive or part of the archive metadata.
So, for example, if we want to take a tar file, modify the
postgresql.auto.conf file it contains, and the gzip the result
and write it out, we can use a bbstreamer_tar_parser to parse the
tar file received from the server, a bbstreamer_recovery_injector
to modify the contents of postgresql.auto.conf, a
bbstreamer_tar_archiver to replace the tar headers for the file
modified in the previous step with newly-built ones that are
correct for the modified file, and a bbstreamer_gzip_writer to
gzip and write the resulting data. Only the objects with "tar"
in the name know anything about the tar archive format, and in
theory we could re-archive using some other format rather than
"tar" if somebody wanted to write the code.
These chances do add a substantial amount of code, but I think the
result is a lot more maintainable and extensible. pg_basebackup.c
itself shrinks by roughly a third, with a lot of the complexity
previously contained there moving into the newly-added files.
Patch by me. The larger patch series of which this is a part has been
reviewed and tested at various times by Andres Freund, Sumanta
Mukherjee, Dilip Kumar, Suraj Kharage, Dipesh Pandit, Tushar Ahuja,
Mark Dilger, Sergei Kornilov, and Jeevan Ladhe.
Discussion: https://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
Discussion: https://postgr.es/m/CA+TgmoZvqk7UuzxsX1xjJRmMGkqoUGYTZLDCH8SmU1xTPr1Xig@mail.gmail.com
2021-11-05 10:22:07 -04:00
|
|
|
|
|
|
|
|
pfree(mystreamer->basepath);
|
|
|
|
|
pfree(mystreamer);
|
|
|
|
|
}
|