From: | Raghavendra <raghavendra(dot)rao(at)enterprisedb(dot)com> |
---|---|
To: | Bhushan Pathak <bhushan(dot)pathak02(at)gmail(dot)com> |
Cc: | pgsql-general(at)postgresql(dot)org |
Subject: | Re: PostgreSQL 9.2.4 using large amount of memory |
Date: | 2014-01-29 02:19:02 |
Message-ID: | CA+h6AhgcNk+tJBQ8=TycQb8Lz5Yf-YkxJ_0nnnSv+PWhrz-N8A@mail.gmail.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
Thanks
On 27 Jan 2014 22:35, "Bhushan Pathak" <bhushan(dot)pathak02(at)gmail(dot)com> wrote:
>
> Hello,
>
> We have recently shifted to postgresql version 9.2.4 from 9.1.3. After
the migration, we observed that some of our delete queries on single table
[which have triggers, which in turn call other functions] have started
consuming large amounts of memory.
>
> In 9.1.3, this usage was upto 25MB with the same load on the same server.
With 9.2.4 it has jumped upto ~580 MB. We are monitoring the RES column
from top output to get the memory usage.
>
> Our migration method from 9.1.3 to 9.2.4 was take a dump, un-install
9.1.3, install 9.2.4 & restore the dump.
>
> I also went through the thread -
>
http://postgresql.1045698.n5.nabble.com/Memory-usage-after-upgrade-to-9-2-4-td5752733.html
>
> In the thread in the end it is mentioned that there was some data
corruption & points to 9.1.6 release notes. I went through the release
notes & only thing of note that I found was the re-indexing or performing
vacuum operation in case of in-place upgrade, which is not the case for me.
>
> Any help/pointers in debugging would be helpful.
>
> Thanks
> Bhushan
>
Just wanted to know, after upgrade as a part of process have you performed
ANALYZE on the database. I agree this might not relate to the question but
am curious to know this issue raised after proper upgrade method.
--Raghav
From | Date | Subject | |
---|---|---|---|
Next Message | Chris Travers | 2014-01-29 02:39:48 | Re: Composite type |
Previous Message | Andrew Sullivan | 2014-01-28 23:55:12 | Re: PostgreSQL specific datatypes very confusing for beginners who use wrappers around JDBC |