On 2025-04-01 01:12, Fujii Masao wrote:
On 2025/03/31 22:45, Yugo Nagata wrote:
I prefer this approach clarifying that consistency and subtransaction
overflow
are separate concepts in the documentation.
Here are minor comments on the patch:
Thanks for the review!
- case CAC_NOTCONSISTENT:
- if (EnableHotStandby)
+ case CAC_NOTHOTSTANDBY:
+ if (!EnableHotStandby)
ereport(FATAL,
(errcode(ERRCODE_CANNOT_CONNECT_NOW),
errmsg("the database system is not yet
accepting connections"),
- errdetail("Consistent recovery state has
not been yet reached.")));
+ errdetail("Hot standby mode is
disabled.")));
+ else if (reachedConsistency)
+ ereport(FATAL,
+ (errcode(ERRCODE_CANNOT_CONNECT_NOW),
+ errmsg("the database system is not
accepting connections"),
+ errdetail("Recovery snapshot is not yet
ready for hot standby."),
+ errhint("To enable hot standby, close
write transactions with more than %d subtransactions on the primary
server.",
+ PGPROC_MAX_CACHED_SUBXIDS)));
else
ereport(FATAL,
(errcode(ERRCODE_CANNOT_CONNECT_NOW),
errmsg("the database system is not
accepting connections"),
- errdetail("Hot standby mode is
disabled.")));
+ errdetail("Consistent recovery state has
not been yet reached.")));
The message says "the database system is not yet accepting
connections" when "Hot standby mode is disabled".
I think "yet" is not necessary in this case. Otherwise, when "Recovery
snapshot is not yet ready for hot standby"
or "Consistent recovery state has not been yet reached", it seems
better to use "yet"
I may have unintentionally modified the error message.
I fixed the patch as suggested. Please check the latest patch
I posted earlier in response to Torikoshi-san.
Thank you for updating the patch!
LGTM.
I feel like changing the status to 'Ready for Committer', but since
Nagata-san may have additional comments, I'm leaving it as 'Needs
Review'.
--
Regards,
--
Atsushi Torikoshi
Seconded from NTT DATA GROUP CORPORATION to SRA OSS K.K.
