Re: BUG #18624: Memory Leak Issue with PostgreSQL Connection During COPY Command Execution. - Mailing list pgsql-bugs

From Tomas Vondra
Subject Re: BUG #18624: Memory Leak Issue with PostgreSQL Connection During COPY Command Execution.
Date
Msg-id e37eb66f-be10-49fa-95dc-253dea622685@vondra.me
Whole thread Raw
In response to Re: BUG #18624: Memory Leak Issue with PostgreSQL Connection During COPY Command Execution.  (Abhishek Mittal <abhmittal@gmail.com>)
List pgsql-bugs
On 10/3/24 11:49, Abhishek Mittal wrote:
> Hi,
> 
> I'm currently facing a blocker in my project. Could we please check this
> issue.
> 
> Thank you for your assistance.
> 

Please, don't top post. Reply in-line.

You have not reported if you tried upgrading to 12.20. Chances are this
issue was already fixed since 12.16, people are unlikely to spend time
investigating this, only to find it was a waste of time.

Also, Tom asked you a couple important questions you ignored:

> If updating to 12.20 does not resolve the problem, then we are
> going to need more information.  Is this straight community
> Postgres, or some variant?  If you built it yourself, what
> options did you use?  Do you have any extensions installed?

Please provide this info. It was suggested to try with "jit=off", so
maybe try that too - I personally don't see why would that matter for
the COPY command, but maybe it does ...

Finally, I wonder if you're simply misinterpreting what RES means. It's
not "private memory used by the process". So if you're simply seeing RES
to grow to 10GB and conclude "memory leak", it may not be the case. It
could be shared memory, for example - you need to inspect "smaps" for
the process to see what exactly the RES figure includes.


regards

-- 
Tomas Vondra




pgsql-bugs by date:

Previous
From: Muralikrishna Bandaru
Date:
Subject: Re: BUG #18615: installer cannot be executed as "nt-autorität\system"
Next
From: Dean Rasheed
Date:
Subject: Re: BUG #18634: Wrong varnullingrels with merge ... when not matched by source