From 5f4e53e5a2d95485095ef69b3a1ed732d7888f1e Mon Sep 17 00:00:00 2001 From: Jason Peng <86845444+jpenghashi@users.noreply.github.com> Date: Mon, 22 Apr 2024 15:45:20 -0700 Subject: [PATCH] Update health.mdx (#26264) To address the confusion on 472 status code, in addition to explain in rare occasions when a 429 could be returned. --- website/content/api-docs/system/health.mdx | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/website/content/api-docs/system/health.mdx b/website/content/api-docs/system/health.mdx index 8bc1b94ec7..137ba2329b 100644 --- a/website/content/api-docs/system/health.mdx +++ b/website/content/api-docs/system/health.mdx @@ -25,11 +25,15 @@ The default status codes are: - `200` if initialized, unsealed, and active - `429` if unsealed and standby -- `472` if disaster recovery mode replication secondary and active +- `472` if disaster recovery secondary (both active and standby nodes within the DR secondary) - `473` if performance standby - `501` if not initialized - `503` if sealed + +In rare occasions such as during cluster instability, a node may return 429 even when it was part of a DR secondary (472) or a perf-standby (473). When configuring a Load Balancer based on health status API, it's important to recognize that a 429 indicates a standby node that doesn't process the request itself, read or write. + + ### Parameters - `standbyok` `(bool: false)` – Specifies if being a standby should still return