public inbox for [email protected]  
help / color / mirror / Atom feed
From: Jeremy Finzel <[email protected]>
To: Scott Marlowe <[email protected]>
Cc: Jerry Sievers <[email protected]>
Cc: Peter Geoghegan <[email protected]>
Cc: postgres performance list <[email protected]>
Subject: Re: Odd sudden performance degradation related to temp object churn
Date: Fri, 18 Aug 2017 09:21:36 -0500
Message-ID: <CAMa1XUg1MErrJifeeTNse+TQdbEZjs3t+tzOvnu7KCW9tUqcsA@mail.gmail.com> (raw)
In-Reply-To: <CAOR=d=3+N1467oddeN3E7r2PKELQt+85s+c+JjP-tgRHoYukpA@mail.gmail.com>
References: <CAMa1XUi6pbugdB7zUA6weoPjmUHTzNLZcSEAh06-7FZRzkuHQw@mail.gmail.com>
	<CAH2-WznKCt_UBaPEY=WwBof+TkTaDYHbZmDsFE=Vigy3Rqi35A@mail.gmail.com>
	<[email protected]>
	<CAOR=d=3+N1467oddeN3E7r2PKELQt+85s+c+JjP-tgRHoYukpA@mail.gmail.com>
List-Unsubscribe:  <mailto:[email protected]?body=unsub%20pgsql-performance>

On Tue, Aug 15, 2017 at 12:07 PM, Scott Marlowe <[email protected]>
wrote:

> So do iostat or iotop show you if / where your disks are working
> hardest? Or is this CPU overhead that's killing performance?
>

Sorry for the delayed reply.  I took a look in more detail at the query
plans from our problem query during this incident.  There are actually 6
plans, because there were 6 unique queries.  I traced one query through our
logs, and found something really interesting.  That is that all of the
first 5 queries are creating temp tables, and all of them took upwards of
500ms each to run.  The final query, however, is a simple select from the
last temp table, and that query took 0.035ms!  This really confirms that
somehow, the issue had to do with *writing *to the SAN, I think.  Of course
this doesn't answer a whole lot, because we had no other apparent issues
with write performance at all.

I also provide some graphs below.

7pm-3am on 8/10 (first incidents were around 10:30pm, other incidents ~1am,
2am):

Local Disk IO:

[image: Screen Shot 2017-08-18 at 8.20.06 AM.png]

SAN IO:

[image: Screen Shot 2017-08-18 at 8.16.59 AM.png]

CPU:

[image: Screen Shot 2017-08-18 at 8.20.58 AM.png]

7-9pm on 8/10 (controlled attempts starting a little after 7):

CPU:

[image: Screen Shot 2017-08-18 at 8.43.35 AM.png]

Write IO on SAN:

[image: Screen Shot 2017-08-18 at 8.44.32 AM.png]

Read IO on Local disk:

[image: Screen Shot 2017-08-18 at 8.46.27 AM.png]

Write IO on Local disk:

[image: Screen Shot 2017-08-18 at 8.46.58 AM.png]


reply

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Reply to all the recipients using the --to and --cc options:
  reply via email

  To: [email protected]
  Cc: [email protected], [email protected], [email protected], [email protected]
  Subject: Re: Odd sudden performance degradation related to temp object churn
  In-Reply-To: <CAMa1XUg1MErrJifeeTNse+TQdbEZjs3t+tzOvnu7KCW9tUqcsA@mail.gmail.com>

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

This inbox is served by agora; see mirroring instructions
for how to clone and mirror all data and code used for this inbox