I don't know Postgres that well either, but in general, when you have a concurrency issue, you use a mutex or a semaphore to fence that area.
On Fri, Jan 31, 2014 at 6:58 AM, Alan Ward <[email protected]> wrote: > Hi Michael, > Thanks for your quick response. > > I agree it is not a 0MQ issue, I am just trying to find out if anybody else > had the same problem in this environment and solved it. > > Thanks > Alan Ward > > (Sent from my iPad) > >> On 31 Jan 2014, at 12:06, Michael Powell <[email protected]> wrote: >> >> This doesn't sound like a 0MQ issue, but I don't know the 0MQ >> threading model that well. I sounds more like a Postgres issue. You >> want to protect your database with a memory fence like a semaphore or >> mutex. Or rethink your architecture. >> >>> On Fri, Jan 31, 2014 at 5:47 AM, Alan Ward <[email protected]> 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 >>> [email protected] >>> http://lists.zeromq.org/mailman/listinfo/zeromq-dev >> _______________________________________________ >> zeromq-dev mailing list >> [email protected] >> http://lists.zeromq.org/mailman/listinfo/zeromq-dev > _______________________________________________ > zeromq-dev mailing list > [email protected] > http://lists.zeromq.org/mailman/listinfo/zeromq-dev _______________________________________________ zeromq-dev mailing list [email protected] http://lists.zeromq.org/mailman/listinfo/zeromq-dev
