I have tested it and I can now confirm that the issue is with pg8000. I think its a good idea to mention this in the WIKI.
On Friday, July 29, 2016 at 12:12:07 AM UTC+5:30, Niphlod wrote: > > could you please try with psycopg2 ? > > On Thursday, July 28, 2016 at 6:53:10 PM UTC+2, Abhishek Ram wrote: >> >> I am using the default driver pg8000 >> >> On Thursday, July 28, 2016 at 7:47:41 PM UTC+5:30, Niphlod wrote: >>> >>> what driver are you using ? it's rather strange that postrgresql acts up >>> and eat all the available memory, if you didn't tinker with the defaults >>> settings of postgresql... >>> >>> On Wednesday, July 27, 2016 at 5:09:34 AM UTC+2, Abhishek Ram wrote: >>>> >>>> Well the memory leak was from the postgres processes spawned by the >>>> web2py scheduler process. I am sure about this as I have seen TOP and >>>> confirmed that it was these processes that were consuming the memory. I >>>> even tried with a basic app and the postgres processes continued to >>>> consume >>>> memory. So finally I shifted to mysql and memory consumption is now stable. >>>> >>>> On Tuesday, July 19, 2016 at 12:24:27 AM UTC+5:30, Niphlod wrote: >>>>> >>>>> you're seeing a totally normal thing: postgresql spawns a different >>>>> process for each connection. >>>>> Those processes in linux are fork()s, so even if you SEEM to notice a >>>>> skyrocket in memory utilization summing all different processes (e.g. in >>>>> "top"), in reality it's not sucking up every bit of RAM of your server. >>>>> >>>>> I'm sure this is not your case, but if you see postgresql processes >>>>> comsuming memory, the leak in postgresql processes CAN'T be generated by >>>>> web2py in any circumnstance. Unless you tinkered a lot with >>>>> postgresql.conf >>>>> (which is pretty conservative in all distros) you won't ever see >>>>> postgresql >>>>> leaking. >>>>> >>>>> On Sunday, July 17, 2016 at 5:50:04 AM UTC+2, Abhishek Ram wrote: >>>>>> >>>>>> Hi, >>>>>> >>>>>> I am using the web2py scheduler to run some background jobs. I am >>>>>> starting the scheduler using systemd and once started it also spawns two >>>>>> postgres processes as that is my DB. Now the problem is that the >>>>>> postgres >>>>>> processes continue to consume more and more memory even when no process >>>>>> is >>>>>> running. >>>>>> >>>>>> So my question is what is causing the scheduler process to do this? >>>>>> >>>>>> P.S. The postgres processes linked to the main web2py process do not >>>>>> exhibit this behaviour. >>>>>> >>>>>> Regards >>>>>> Abhishek Ram >>>>>> >>>>> -- Resources: - http://web2py.com - http://web2py.com/book (Documentation) - http://github.com/web2py/web2py (Source code) - https://code.google.com/p/web2py/issues/list (Report Issues) --- You received this message because you are subscribed to the Google Groups "web2py-users" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/d/optout.
