[pgpool-hackers: 4271] Removing life check in watchdog

[Tatsuo Ishii]

We have life check process to detect failures of watchdog and/or
communication between each watchdog node. On the other hand we have a
communication path between watchdog nodes. The latter is used for
communication (e.g. leader election) but also is used for node down
detection. So I think there's functional overlapping among the life
check process and watchdog communication. This leads me to an extreme
idea: why can't we remove the life check process?

So I created a modified pgpool so that it does not start the life
check process. Also I modified
028.watchdog_enable_consensus_with_half_votes because it relies on
life check starting log to detect whether watchdog system is usable
after pgpool restart (see attached patches).

Then all regression tests passed! I am not saying that we can
immediately remove the life check process. We need more tests of
course. But I believe this is worth a discussion with pgpool
developers toward Pgpool-II 4.5.

Best reagards,
--
Tatsuo Ishii
SRA OSS LLC
English: http://www.sraoss.co.jp/index_en/
Japanese:http://www.sraoss.co.jp
-------------- next part --------------
A non-text attachment was scrubbed...
Name: removing_life_check.patch
Type: text/x-patch
Size: 968 bytes
Desc: not available
URL: <http://www.pgpool.net/pipermail/pgpool-hackers/attachments/20230124/2da81bdf/attachment.bin>