Re: Measuring replay lag

From: Thomas Munro <thomas(dot)munro(at)enterprisedb(dot)com>
To: Masahiko Sawada <sawada(dot)mshk(at)gmail(dot)com>
Cc: Pg Hackers <pgsql-hackers(at)postgresql(dot)org>
Subject: Re: Measuring replay lag
Date: 2016-11-22 09:27:14
Message-ID: CAEepm=0vv_YetnKEheCYULQU6RT3yZonVmeioCr_mad7a=BeYQ@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox
Thread:
Lists: pgsql-hackers

On Tue, Nov 8, 2016 at 2:35 PM, Masahiko Sawada <sawada(dot)mshk(at)gmail(dot)com> wrote:
> replay_lag_sample_interval is 1s by default but I got 1000s by SHOW command.
> postgres(1:36789)=# show replay_lag_sample_interval ;
> replay_lag_sample_interval
> ----------------------------
> 1000s
> (1 row)

Oops, fixed.

>> 1. The lag is measured in time, not LSN difference.
>> 2. The lag time is computed using two observations of a single
>> server's clock, so there is no clock skew.
>> 3. The lag is updated even between commits (during large data loads etc).
>
> I agree with this approach.

Thanks for the feedback.

> I think that you need to change sendFeedback() in pg_recvlogical.c and
> receivexlog.c as well.

Right, fixed.

Thanks very much for testing! New version attached. I will add this
to the next CF.

--
Thomas Munro
http://www.enterprisedb.com

Attachment Content-Type Size
replay-lag-v13.patch application/octet-stream 30.9 KB

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Kyotaro HORIGUCHI 2016-11-22 09:27:49 Re: Fix checkpoint skip logic on idle systems by tracking LSN progress
Previous Message Amit Langote 2016-11-22 09:15:38 Re: Declarative partitioning - another take