[pgpool-general: 5910] Re: Fwd: repeat the check of the trusted servers

[Muhammad Usama]

Hi

Sorry for the delayed response, Somehow my email filters are messed up and
community mails are not properly
showing in my inbox.
It seems that the log messages you shared is from the Pgpool server that
did not suffer the connectivity problem. Can you also share
the pgpool logs from the pgpool server that shutdown itself because of ping
to upstream failure.

Thanks
Best Regards
Muhammad Usama




On Mon, Dec 18, 2017 at 8:21 PM, Giancarlo Celli <giancarlocelli at gmail.com>
wrote:

> So I'm sharing this mail on the pgpool-general, as requested.
> My pgpool version number is: 3.6.6
> My postgresql versione is: 9.6.3
> Can someone help me, please?
>
>
> ---------- Forwarded message ----------
> From: Tatsuo Ishii <ishii at sraoss.co.jp>
> Date: 2017-12-15 23:04 GMT+01:00
> Subject: Re: [pgpool-general: 5829] repeat the check of the trusted servers
> To: giancarlocelli at gmail.com
> Cc: ishii at sraoss.co.jp
>
>
> Ok, from the log I guess it maybe related to watchdog part. Usama is
> the authority of watchdog, and I suggest you share this mail on the
> pgpool-general mailing list so that he could join the discussion
> unless you have some reasons to not want to do that. And recently
> watchdog has been fixed and enhanced. It will be helpful if you
> provide accurate version number of Pgpool-II.
>
> Best regards,
> --
> Tatsuo Ishii
> SRA OSS, Inc. Japan
> English: http://www.sraoss.co.jp/index_en.php
> Japanese:http://www.sraoss.co.jp
>
> From: Giancarlo Celli <giancarlocelli at gmail.com>
> Subject: Re: [pgpool-general: 5829] repeat the check of the trusted servers
> Date: Fri, 15 Dec 2017 17:35:17 +0100
> Message-ID: <CAD57xQ63HtgeD5Pz69DKn59qP2dvcQibK8TXgZcxXThH9ND5mw at mail.gm
> ail.com>
>
> > Hi,
> > no, this is a dedicated server.
> > The strange thing is that the provider does not show connection problems,
> > as you can see in the attached image (green: input data; blue: output
> data).
> > Instead, the following lines appear in the pgpool log (I've hidden part
> of
> > the IP for security reasons):
> >
> > 2017-12-15 03:44:45: pid 2319: LOG: read from socket failed, remote end
> > closed the connection
> > 2017-12-15 03:44:45: pid 2319: LOG: client socket of 95.110.x.1:5432
> Linux
> > standby1 is closed
> > 2017-12-15 03:44:45: pid 2319: LOG: remote node "95.110.x.1:5432 Linux
> > standby1" is shutting down
> > 2017-12-15 03:44:45: pid 2319: LOG: removing watchdog node
> "95.110.x.1:5432
> > Linux standby1" from the standby list
> > 2017-12-15 03:44:45: pid 2319: LOG: new watchdog node connection is
> > received from "95.110.x.1:25742"
> > 2017-12-15 03:44:45: pid 2319: LOG: new node joined the cluster
> > hostname:"95.110.x.1" port:9000 pgpool_port:5432
> > 2017-12-15 03:44:45: pid 2319: LOG: new outbond connection to
> > 95.110.x.1:9000
> > 2017-12-15 03:44:45: pid 2319: LOG: remote node "95.110.x.1:5432 Linux
> > standby1" is shutting down
> > 2017-12-15 03:44:56: pid 2321: LOG: informing the node status change to
> > watchdog
> > 2017-12-15 03:44:56: pid 2321: DETAIL: node id :1 status = "NODE DEAD"
> > message:"No heartbeat signal from node"
> > 2017-12-15 03:44:56: pid 2319: LOG: new IPC connection received
> > 2017-12-15 03:44:56: pid 2319: LOG: received node status change ipc
> message
> > 2017-12-15 03:44:56: pid 2319: DETAIL: No heartbeat signal from node
> > 2017-12-15 03:44:56: pid 2319: LOG: remote node "95.110.x.1:5432 Linux
> > standby1" is shutting down
> >
> > In the log I hid part of the IP for security reasons.
> >
> > Do you have any suggestions?
> >
> > 2017-12-15 16:14 GMT+01:00 Tatsuo Ishii <ishii at sraoss.co.jp>:
> >
> >> > Hi
> >> > I've one standby server who has some connectivity problems and it
> loses
> >> its
> >> > connection for short periods. When this happens it fails to ping
> trusted
> >> > servers and so pgpool process was stopped. Is there a specific
> parameter
> >> > that can be modified to make sure that the check on trusted servers is
> >> > repeated several times with an interval of a few seconds before the
> >> pgpool
> >> > is stopped?
> >>
> >> Unfortunately there is no such a retry feature for trusted
> >> servers.
> >>
> >> Just out of curiosity, is your network running on a cloud service? If
> >> it's a common phenomena on cloud services, then probably we should
> >> implement the retry feature for trusted servers.
> >>
> >> Best regards,
> >> --
> >> Tatsuo Ishii
> >> SRA OSS, Inc. Japan
> >> English: http://www.sraoss.co.jp/index_en.php
> >> Japanese:http://www.sraoss.co.jp
> >>
>
>
> _______________________________________________
> pgpool-general mailing list
> pgpool-general at pgpool.net
> http://www.pgpool.net/mailman/listinfo/pgpool-general
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.sraoss.jp/pipermail/pgpool-general/attachments/20180201/e2fa5556/attachment.html>