[prev in list] [next in list] [prev in thread] [next in thread] 

List:       zeromq-dev
Subject:    Re: [zeromq-dev] Question about CZMQ threading and Postgres
From:       "Alan Ward" <alan.ward () btinternet ! com>
Date:       2014-01-31 15:35:09
Message-ID: 001401cf1e9a$06c53ca0$144fb5e0$ () btinternet ! com
[Download RAW message or body]

Hi Pieter,
Thanks for the insight.

On reflection, it is unreasonable of me to expect Postgres client to operate
across multiple threads within one process. The failure mode supports that -
as it was complaining that the database was already disconnected , when it
tried to disconnect. Presumably using the same static data within the
process.

I'll move to using different processes: just need to manage the connection
pool as I want the number of processes to be configurable.

Thanks,
Alan

-----Original Message-----
From: zeromq-dev-bounces@lists.zeromq.org
[mailto:zeromq-dev-bounces@lists.zeromq.org] On Behalf Of Pieter Hintjens
Sent: 31 January 2014 13:18
To: ZeroMQ development list
Subject: Re: [zeromq-dev] Question about CZMQ threading and Postgres

Hi Alan,

zthread_fork and zthread_new will look the same to Postgres.

The difference is for the CZMQ infrastructure; the first creates an
"attached thread" (sharing the same context and with a pipe back to the
parent), while the second creates a "detached thread" (shares nothing, so
acts more like a separate process, and cannot talk over inproc to the
parent).

Could you post a *minimal* piece of code that provokes the error so we can
look at it?

Could be the Postgres client has static variables and simply can't be called
from 2+ threads in the same process. In that case, it is trivial to solve.
Create a single Postgres client thread (attached :-) and talk to it over
inproc.

-Pieter

On Fri, Jan 31, 2014 at 12:47 PM, Alan Ward <alan.ward@btinternet.com>
wrote:
> Hi,
> I would like to find out if anyone has successfully run multiple 
> worker threads created by CZMQ zthread_fork, where each thread is the 
> same and running Postgres updates to the same database.
>
> My system works fine if I use one worker thread, but if I increase it 
> to
> (say) 3, Postgres fails with (undocumented) YE000 errors.
>
> Searching the net finds some instances of this error being reported in 
> 2007 and it relates to problems with concurrent access from multiple
threads.
>
> Should I be using zthread_new?
>
> Thanks,
> Alan
>
>
> _______________________________________________
> zeromq-dev mailing list
> zeromq-dev@lists.zeromq.org
> http://lists.zeromq.org/mailman/listinfo/zeromq-dev
_______________________________________________
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev


-----
No virus found in this message.
Checked by AVG - www.avg.com
Version: 2014.0.4259 / Virus Database: 3684/7047 - Release Date: 01/30/14

_______________________________________________
zeromq-dev mailing list
zeromq-dev@lists.zeromq.org
http://lists.zeromq.org/mailman/listinfo/zeromq-dev
[prev in list] [next in list] [prev in thread] [next in thread]