[pgpool-general: 6711] Re: pgpool often goes down on a node

[Tatsuo Ishii]

> Hi,
> I answer you point by point:
> 
> On Thursday, September 12, 2019, Tatsuo Ishii <ishii at sraoss.co.jp>
> wrote:
>>>>> Hi,
>>>>> on a node that had never had problems, the pgpool often goes down from
>>>>> about 2 weeks. In the log it looks like a connection problem that
>>>>> really doesn't seem to exist. When I restart pgpool after a few hours
>>>>> it goes down again. What could be the problem?
>>>>> The log also seems to fail to ping trusted servers (which are google
>>>>> ones).
>>>>> An extract of pgpool.log is in attached file.
>>>>> Please, help me.
>>>>> Thanks!
>>>>> Best regards.
>>>>> Giancarlo
>>>> "Goes down" means clients cannot connect to the pgpool or any other
>>>> situation?
> It means that pgpool process goes down.

Oh, ok. That's an expected behavior. If pgpool cannot reach to non of
trusted servers, it will suicide.
> 
>>>>> In the log it looks like a connection problem that
>>>>> really doesn't seem to exist. When I restart pgpool after a few hours
>>>> Can you please point out which line number in the log file you are
>>>> referring to when you say "looks like a connection problem"?
> because ping fails to all trusted servers:
> 
> 2019-09-11 00:17:21: pid 26292: WARNING: watchdog failed to ping
> host"8.8.8.8"
> 2019-09-11 00:17:21: pid 26292: DETAIL: ping process exits with code:
> 1
> 2019-09-11 00:17:21: pid 26292: WARNING: watchdog failed to ping
> host"8.8.4.4"
> 2019-09-11 00:17:21: pid 26292: DETAIL: ping process exits with code:
> 1
> 2019-09-11 00:17:21: pid 26292: WARNING: watchdog lifecheck, failed to
> connect to any trusted servers
> 
> But really the connection does not go down.
> Maybe a watchdog problem?
> Is it possible to set up a number of ping attempts before throwing
> down pgpools?

There's no such a feature in pgpool for now.

Best regards,
--
Tatsuo Ishii
SRA OSS, Inc. Japan
English: http://www.sraoss.co.jp/index_en.php
Japanese:http://www.sraoss.co.jp