Re: checkpoint and recovering process use too much memory

From: tao tony <tonytao0505(at)outlook(dot)com>
To: Justin Pryzby <pryzby(at)telsasoft(dot)com>
Cc: "pgsql-general(at)postgresql(dot)org" <pgsql-general(at)postgresql(dot)org>
Subject: Re: checkpoint and recovering process use too much memory
Date: 2017-11-06 01:11:35
Message-ID: MWHPR13MB138955AB55FAAE0004A5AC5CAA500@MWHPR13MB1389.namprd13.prod.outlook.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-general

Thank you,  Justin Pryzby.
I reset shared_buffer to 16GB,and the  memory usage of  checkpoint and
recovering just stayed at 16GB.

  PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+ COMMAND
192956 postgres  20   0 18.5g  16g  16g S  1.3 25.9  19:44.69 postgres:
startup process   recovering 00000004000008A300000035
192960 postgres  20   0 18.5g  16g  16g S  0.7 25.8  11:13.79 postgres:
checkpointer process
192951 postgres  20   0 18.5g 1.9g 1.9g S  0.0  3.1   0:01.75
/usr/pgsql-9.6/bin/postmaster -D /data/pgdata

Thank you again for your help.

On 11/03/2017 10:21 AM, Justin Pryzby wrote:
> On Fri, Nov 03, 2017 at 01:43:32AM +0000, tao tony wrote:
>> I had an asynchronous steaming replication HA cluster.Each node had 64G memory.pg is 9.6.2 and deployed on centos 6.
>>
>> Last month the database was killed by OS kernel for OOM,the checkpoint process was killed.
> If you still have logs, was it killed during a large query? Perhaps one using
> a hash aggregate?
>
>> I noticed checkpoint process occupied memory for more than 20GB,and it was growing everyday.In the hot-standby node,the recovering process occupied memory as big as checkpoint process.
> "resident" RAM of a postgres subprocess is often just be the fraction of
> shared_buffers it's read/written. checkpointer must necessarily read all dirty
> pages from s-b and write out to disk (by way of page cache), so that's why its
> RSS is nearly 32GB. And the recovery process is continuously writing into s-b.
>
>> Now In the standby node,checkpoint and recovering process used more then 50GB memory as below,and I worried someday the cluster would be killed by OS again.
>>
>> PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
>> 167158 postgres 20 0 34.9g 25g 25g S 0.0 40.4 46:36.86 postgres: startup process recovering 00000004000008550000004B
>> 167162 postgres 20 0 34.9g 25g 25g S 0.0 40.2 17:58.38 postgres: checkpointer process
>>
>> shared_buffers = 32GB
> Also, what is work_mem ?
>
> Justin

In response to

Browse pgsql-general by date

  From Date Subject
Next Message hmidi slim 2017-11-06 08:17:25 Combine multiple text search configuration
Previous Message Bret Stern 2017-11-05 21:39:34 Query Improvement??