Howard Chu
|
d1221975d6
|
ITS#5809 fewer false detections of renames
|
2008-11-14 23:48:38 +00:00 |
|
Howard Chu
|
51c87bb9c7
|
ITS#5809 renames are not exclusive from modifies
|
2008-11-13 22:46:49 +00:00 |
|
Howard Chu
|
4d51be0c9b
|
ITS#5798, don't use per-DB schema-check flag
|
2008-11-08 16:48:48 +00:00 |
|
Howard Chu
|
5dc58028c1
|
ITS#5710 fix stupid updateCookie hack, use explicit o_dont_replicate flag
|
2008-11-06 21:10:25 +00:00 |
|
Howard Chu
|
fa245f7491
|
ITS#5781 fix modification of MV attrs with no EQ rule
|
2008-10-30 18:22:11 +00:00 |
|
Hallvard Furuseth
|
db5943fbd6
|
ITS#4467: Fix snprintf/WHATSLEFT buffer overflow tests.
Also make WHATSLEFT unsigned.
|
2008-10-24 13:16:06 +00:00 |
|
Pierangelo Masarati
|
faddae3a0e
|
check for overflows
|
2008-10-22 23:37:34 +00:00 |
|
Hallvard Furuseth
|
7b3bdf2185
|
Warning cleanup: signed meets unsigned. ber_flatten2() returns -1 on
error, not LBER_ERROR.
|
2008-10-13 10:18:15 +00:00 |
|
Pierangelo Masarati
|
6845c2510d
|
don't free and reuse filter structure; don't rewrite undefined filter values (ITS#5731,ITS#5732)
|
2008-10-09 22:34:43 +00:00 |
|
Rein Tollevik
|
f1f9a6280a
|
ITS#5726: Revert 1.392, the runqueue must always be inspected.
|
2008-10-05 22:35:44 +00:00 |
|
Rein Tollevik
|
c5f584e62b
|
Revert 1.410, i.e back to lock again. ITS#5725 fixes the lock situation.
|
2008-10-05 19:29:00 +00:00 |
|
Rein Tollevik
|
fd190659a7
|
ITS#5719 Reset interval after rescheduling a paused task.
|
2008-09-29 20:01:40 +00:00 |
|
Howard Chu
|
a49d43ca27
|
Revert 1.394, see ITS#5454. It still locks up without a trylock.
|
2008-09-25 10:31:49 +00:00 |
|
Pierangelo Masarati
|
f4b98b256c
|
use the most recent contextCSN (ITS#5675)
|
2008-09-02 20:35:51 +00:00 |
|
Howard Chu
|
d4216935d7
|
ITS#5618 log Search result errors, force Deref to Never
|
2008-08-09 13:15:54 +00:00 |
|
Rein Tollevik
|
a42c7f96f5
|
ITS#5600 Replace objectClass, delete/add can fail if superior is added.
|
2008-07-08 13:55:09 +00:00 |
|
Rein Tollevik
|
f94cd4b518
|
ITS#5597 fix CSN compare when provider has CSNs with other sids.
|
2008-07-06 14:15:42 +00:00 |
|
Rein Tollevik
|
99d004189a
|
ITS#5596 contextCSN updates are internal operations.
|
2008-07-04 23:44:39 +00:00 |
|
Howard Chu
|
4b2c003907
|
More for ITS#5536
|
2008-05-29 23:31:05 +00:00 |
|
Howard Chu
|
5a97371ef3
|
ITS#5536 send SID for glued DBs as well as multimaster
|
2008-05-29 01:46:19 +00:00 |
|
Howard Chu
|
61a70b4f96
|
ITS#5403 fix sasl_ssf retrieval on 64 bit Solaris
|
2008-05-28 01:43:05 +00:00 |
|
Howard Chu
|
b2a0eb7075
|
Fix prev commit
|
2008-05-01 23:56:01 +00:00 |
|
Howard Chu
|
b94eac4778
|
Fix prev commit (again) - set in wrong order
|
2008-04-30 18:35:02 +00:00 |
|
Howard Chu
|
bdd493380d
|
Fix prev commit
|
2008-04-30 12:52:45 +00:00 |
|
Howard Chu
|
4673c99e96
|
More for ITS#5470, add csns to presence search filter in MMR
|
2008-04-29 01:05:09 +00:00 |
|
Howard Chu
|
fd0699edf4
|
Fix typo in 1.389
|
2008-04-24 10:19:20 +00:00 |
|
Howard Chu
|
1af2849f7c
|
ITS#5470 also ignore presentlist if number of CSNs don't match
|
2008-04-19 23:27:49 +00:00 |
|
Howard Chu
|
26c3fd67b8
|
ITS#5470 ignore presentlist on partial multimaster refreshes
|
2008-04-17 01:42:40 +00:00 |
|
Howard Chu
|
61b5f6d652
|
ITS#5454 regular mutex lock is safe
|
2008-04-11 12:12:06 +00:00 |
|
Howard Chu
|
d9dac0c8bc
|
ITS#5376, #5378 delta-sync should fallback in other cases too
|
2008-04-09 12:01:11 +00:00 |
|
Howard Chu
|
8027d8afb0
|
ITS#5454 proposed fix. May still need to change initial trylock to a
regular mutex_lock.
|
2008-04-05 00:45:59 +00:00 |
|
Howard Chu
|
f2de258e3d
|
ITS#5442 slapd_rq mutexes must always be used
|
2008-04-02 03:08:06 +00:00 |
|
Howard Chu
|
213c9358b8
|
ITS#5430 syncprov/glue can be used two different ways
|
2008-03-20 10:58:00 +00:00 |
|
Howard Chu
|
d3e8df3940
|
ITS#5432 fix compare_csns loop conditions
|
2008-03-20 00:10:31 +00:00 |
|
Howard Chu
|
d711b538ad
|
do_syncrep1 always reads contextCSN from the main DB
|
2008-03-20 00:05:22 +00:00 |
|
Howard Chu
|
f77bd7ff6e
|
ITS#5426 replace the entire contextCSN in updateCookie
|
2008-03-14 18:00:09 +00:00 |
|
Howard Chu
|
e3849e1a9c
|
Fix prev commit, check for NULL syncCookie
|
2008-03-13 07:30:56 +00:00 |
|
Howard Chu
|
b101dab6c6
|
More fixes for syncCookie updating
|
2008-03-13 07:02:26 +00:00 |
|
Howard Chu
|
86c4dca647
|
Partially revert prev, redundant.
|
2008-03-13 05:41:12 +00:00 |
|
Howard Chu
|
af80427965
|
In mirrormode, also ask syncprov for initial cookie. It may have been
cached from a recent update.
|
2008-03-13 05:18:28 +00:00 |
|
Howard Chu
|
013f78090d
|
ITS#5413 fix typo
|
2008-03-12 02:57:49 +00:00 |
|
Howard Chu
|
a287573d2d
|
ITS#5407 more checks for pool pausing
|
2008-03-08 23:51:07 +00:00 |
|
Quanah Gibson-Mount
|
b2b8a3a0bf
|
ITS#5376: Drop permissive modify, it isn't necessary
|
2008-02-13 19:02:57 +00:00 |
|
Howard Chu
|
ec5a7a9576
|
ITS#5376 must abandon current session before starting new refresh
|
2008-02-13 07:37:29 +00:00 |
|
Howard Chu
|
77618fae10
|
ITS#5376 modifies need to be permissive due to refresh/fallback sync
|
2008-02-13 07:09:18 +00:00 |
|
Howard Chu
|
947e3bbe1d
|
ITS#5376 - deltasync must fallback for missing entries
|
2008-02-13 04:36:58 +00:00 |
|
Howard Chu
|
f26657d81c
|
In do_syncrepl, don't wait if a previous syncrepl thread is still running.
|
2008-02-12 18:49:37 +00:00 |
|
Howard Chu
|
99e1adc1f2
|
Cleanup debug msg, check for redundant updates
|
2008-02-10 02:08:08 +00:00 |
|
Howard Chu
|
5cfa1386ea
|
ITS#5362 syncrepl_updateCookie doesn't need to graduate_commit_csn, the
backends already do it.
|
2008-02-10 02:06:31 +00:00 |
|
Kurt Zeilenga
|
c890c96d13
|
Happy New Year (belated)
|
2008-01-08 00:19:56 +00:00 |
|