Re: My Experiment of PG crash when dealing with huge amount of data - Mailing list pgsql-general

From Tom Lane
Subject Re: My Experiment of PG crash when dealing with huge amount of data
Date
Msg-id 16209.1378089427@sss.pgh.pa.us
Whole thread Raw
In response to Re: My Experiment of PG crash when dealing with huge amount of data  (高健 <luckyjackgao@gmail.com>)
List pgsql-general
=?UTF-8?B?6auY5YGl?= <luckyjackgao@gmail.com> writes:
> If data grows rapidly, maybe our customer will use too much memory , Is
> ulimit  command a good idea for PG?

There's no received wisdom saying that it is.  There's a fairly widespread
consensus that disabling OOM kill can be a good idea, but I don't recall
that many people have tried setting specific ulimits on server processes.

Keep in mind that exceeding a ulimit would cause queries to fail outright
(whether the server was under much load or not), versus just getting
slower if the server starts to swap under too much load.  I can imagine
situations where that would be considered a good tradeoff, but it's hardly
right for everyone.

            regards, tom lane


pgsql-general by date:

Previous
From: 高健
Date:
Subject: Re: My Experiment of PG crash when dealing with huge amount of data
Next
From: Pavel Stehule
Date:
Subject: Re: store multiple rows with the SELECT INTO statement