Re: sorry, too many clients already error

From: Sandeep Gupta <gupta(dot)sandeep(at)gmail(dot)com>
To: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
Cc: "pgsql-general(at)postgresql(dot)org" <pgsql-general(at)postgresql(dot)org>
Subject: Re: sorry, too many clients already error
Date: 2017-05-17 15:42:52
Message-ID: CAAywg7swDGsqbCtB7-penhSyZ0jR993NXN=dNcd5S_5nfb=tXQ@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-general

The log_connections helped. I was able
to diagnose the problem. The connections
are getting closed but my setup is actually generating
too many connections.
Thanks for helping with this.

-sandeep

On Tue, May 16, 2017 at 6:53 PM, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> wrote:
> Sandeep Gupta <gupta(dot)sandeep(at)gmail(dot)com> writes:
>> I have set max_connection = 40.
>> The usage is somewhat not typical. It is basically
>> experiment runs that connect to the database
>> and dump results there.
>
>> The experiments connect through JDBC and
>> they close the connection when they are done.
>
>> I can verify that no more than 20 clients/experiments
>> are active at any given time.
>
> It takes nonzero time for a backend process to actually exit after
> the client has closed the connection. (A client that wants to wait
> for backend exit can do so by waiting for EOF on the socket, but
> I doubt that the JDBC driver does so.) If you're spinning through
> sessions rapidly enough, you might just have >20 sessions that are
> still in process of exiting when a new one comes in and gets this
> error.
>
> It'd also be a good thing to see if the client code is sending an
> actual Terminate protocol message ('X') or is just closing the
> socket. If the latter, maybe it isn't really closing because the
> socket is shared with some child process, or something like that?
>
> Turning on log_connections/log_disconnections might help clarify
> what's happening.
>
> regards, tom lane

In response to

Browse pgsql-general by date

  From Date Subject
Next Message Adrian Myers 2017-05-17 16:03:23 PSQL command line print speed
Previous Message Adrian Klaver 2017-05-17 15:15:13 Re: EnterpriseDB installed PostgreSQL 9.6 vs. REPMGR. Round 4 - compilation issues on RHEL 7.2