[pgpool-general: 8997] Re: Fwd: Pgpool High Load Average

[Bo Peng]

Hi,

> Dear Concern,
> Why I am not able to mail it? Anyone please explain when I can send mail ?
> I am looking for that pgpool solution.

I didn't see your email address in pgpool-general member list.
Have you received any emails from pgpool-general?

Let me answer your original question here.

> > The issue is that, suddenly high load average at other pgpool nodes
> > without any reason. 

> > I have also seen that there is a log written in other
> > pgpool nodes (pgpool2, and pgpool3) like below which is not used because
> > VIP at pgpool1.
> >
> > Jan  9 04:04:27 PGPOOL2[8191]: [214-1] 2024-01-09 04:04:27: pid 8191: LOG:
> >  Replication of node:4 is behind 14935184 bytes from the
> >              primary server (node:0)

The logs above are issued by pgpool streaming replication check process.
Streaming replication check process works even if it is not the leader node (pgpool without VIP).

I think first you should check which processes are consuming a lot of CPU.
If PostgreSQL and running on the same server as Pgpool-II, 
PostgreSQL may be consuming a lot of CPU.
Because the replication delay is high, PostgreSQL may be in a high load situation.

> > My pgpool version : pgpool-II version 4.2.2 (chichiriboshi)
> > My PostgreSQL version: PostgreSQL 13
> > max_pool=4
> > num_init_children=2900
> > tps=4000-6000

Also I noticed that "num_init_children" is set to a high value.
Pgpool-II creates 2900 processes when it starts even if it is not the leader node.
Large processes are likely to use CPU resources.

Best regards,

> ---------- Forwarded message ---------
> From: <pgpool-general-owner at pgpool.net>
> Date: Sat, 13 Jan 2024 at 00:26
> Subject: Re: Pgpool High Load Average
> To: <linuxwasi at gmail.com>
> 
> 
> You are not allowed to post to this mailing list, and your message has
> been automatically rejected.  If you think that your messages are
> being rejected in error, contact the mailing list owner at
> pgpool-general-owner at pgpool.net.
> 
> 
> 
> 
> ---------- Forwarded message ----------
> From: Sheikh Wasiu Al Hasib <linuxwasi at gmail.com>
> To: pgpool-general at pgpool.net
> Cc: Bo Peng <pengbo at sraoss.co.jp>
> Bcc:
> Date: Sat, 13 Jan 2024 00:22:24 +0600
> Subject: Re: Pgpool High Load Average
> Dear Concern,
> Any update on this?
> 
> I am confused how whether anyone getting my message or not.
> 
> Thanks
> 
> On Wed, 10 Jan 2024 at 12:51, Sheikh Wasiu Al Hasib <linuxwasi at gmail.com>
> wrote:
> 
> > Dear Concern,
> >
> > The issue is that, suddenly high load average at other pgpool nodes
> > without any reason. I have also seen that there is a log written in other
> > pgpool nodes (pgpool2, and pgpool3) like below which is not used because
> > VIP at pgpool1.
> >
> > My question is why was this lag info sent to other pgpool2 & pgpool3? ,
> > And why pgpool load average suddenly high like 100-200 at pgpool2, pgpool3
> > nodes? where I did not send any traffic.
> >
> >
> > Jan  9 04:04:27 PGPOOL2[8191]: [214-1] 2024-01-09 04:04:27: pid 8191: LOG:
> >  Replication of node:4 is behind 14935184 bytes from the
> >              primary server (node:0)
> > Jan  9 04:04:27 PGPOOL2 pgpool[8191]: [214-2] 2024-01-09 04:04:27: pid
> > 8191: CONTEXT:  while checking replication time lag
> > Jan  9 04:04:37 PGPOOL2 pgpool[8191]: [215-1] 2024-01-09 04:04:37: pid
> > 8191: LOG:  Replication of node:1 is behind 62281296 bytes from the
> >
> > Jan  9 04:04:27 PGPOOL3[8191]: [214-1] 2024-01-09 04:04:27: pid 8191: LOG:
> >  Replication of node:4 is behind 14935184 bytes from the
> >              primary server (node:0)
> > Jan  9 04:04:27 PGPOOL3 pgpool[8191]: [214-2] 2024-01-09 04:04:27: pid
> > 8191: CONTEXT:  while checking replication time lag
> > Jan  9 04:04:37 PGPOOL3 pgpool[8191]: [215-1] 2024-01-09 04:04:37: pid
> > 8191: LOG:  Replication of node:1 is behind 62281296 bytes from the
> >
> > My pgpool version : pgpool-II version 4.2.2 (chichiriboshi)
> > My PostgreSQL version: PostgreSQL 13
> > max_pool=4
> > num_init_children=2900
> > tps=4000-6000
> >
> >
> > I have been facing this issue last 6 months but have not gotten any
> > solution. I did not find any processes that caused this high load average
> > where CPU and Memory usage is very low.
> >
> > Could you please give me any hints for this issue to solve?
> >
> > If there is any other channel to email you or your team you can also share
> > it with me.
> >


-- 
Bo Peng <pengbo at sraoss.co.jp>
SRA OSS LLC
TEL: 03-5979-2701 FAX: 03-5979-2702
URL: https://www.sraoss.co.jp/