📝 docs: note replicas don't fall back to primary on outage

When RedisCacheReadHosts is set, reads are not retried against the
primary if the replicas are unreachable. Document that this, combined
with the default RedisCacheUnreachableBlock=true, means a replica outage
can block traffic while the primary is healthy.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
This commit is contained in:
mhx
2026-07-25 21:24:11 +02:00
co-authored by Claude Opus 4.8
parent 96beb9192e
commit 73e53c7d69
+1
View File
@@ -449,6 +449,7 @@ make run
- []string - []string
- default: [] - default: []
- List of Redis replica hostnames (host:port) to use for read operations. Reads are distributed round-robin across replicas. Falls back to RedisCacheHost when empty. - List of Redis replica hostnames (host:port) to use for read operations. Reads are distributed round-robin across replicas. Falls back to RedisCacheHost when empty.
- Note: when set, reads are not retried against RedisCacheHost (the primary) if the replicas are unreachable. With RedisCacheUnreachableBlock at its default (true), a replica outage will therefore block/delay requests even though the primary is healthy.
- RedisCachePassword - RedisCachePassword
- string - string
- default: "" - default: ""