Strange. In order to get around the issue I converted it over to run on GAE. This seems to work ok. I notice thought that we're averaging about 2 requests per second.
On Feb 10, 2013, at 3:35 PM, Massimo Di Pierro wrote: > I tried your example but I cannot reproduce the problem (I am on Mac, OSX > 10.7.5 and Python 2.7.3). > > Can others reproduce it? > > On Tuesday, 5 February 2013 13:15:45 UTC-6, pumplerod wrote: > I'm reading up on strace now. in the meanwhile, here is the MVC code (not > sure if I can attach a file to google groups)…. > > > Model: > > -Nothing. I have an empty model file. Other than the default settings that > come with a new controller. > Ordinarily I am using a Model and writing to the SQLite db. But this is the > simplest case I could make that > still replicates the problem. > > View: > > -Don't have this either. I'm only ever returning a json object via REST so > there's no need for a view. > > Controller: > You can see some of what normally gets loaded, which I've commented > out. I also tried to see if gc or cache would help. > I've never used these before though so I may have been doing it > incorrectly. > > the RESTful query would look something like: > http://domain.com/memclimb/request.json?jo={"Test":"memory"} > > > "request.py" (controller)... > _______vv______ > @request.restful() > #@cache(request.env.path_info,time_expire=10,cache_model=cache.ram) > def test(): > #import gc > import gluon.contrib.simplejson > response.view = 'generic.'+request.extension > #gc.enable() > > def GET(*args,**vars): > > jsonObject = request.vars.jo > obj = gluon.contrib.simplejson.loads(jsonObject) > return {'message': 'testing Memory.'} > > #gc.collect() > return locals() > _______^^________ > > Only seems to be a problem under high traffic. So I created this python > script and ran 10 instances of it from another machine. > It's common for our site to have >100 trying to hit it at about the same > time, but I can see the memory climb even with only 10. > > "testLoad.py"... > ________vv________ > #!/usr/bin/env python > > import sys > import os > import urllib > > def main(): > for i in range(1000): > result = > urllib.urlopen('https://domain.com/memclimb/request/test.json?jo={"Test":"memory"}') > json = result.read() > print "done…\n\t%s"%json > > > if __name__ == '__main__': > main() > __________^^________ > > > > On Feb 5, 2013, at 8:30 AM, greaneym wrote: > >> Hello, >> In addition sending the model, view, controller code, you can try the >> following. >> >> It looks like dreamhost is using debian linux. You can try using the strace >> command in a shell while your process is running. You can run it against >> different process ids (pid) to help figure out what is associated with the >> problem. >> >> Here is a link that may be useful in running strace. >> http://www.hokstad.com/5-simple-ways-to-troubleshoot-using-strace >> Save the output in a file so you can look at it. >> >> Also you can run "top" in a shell to see what pid, program is running more >> than others and that may help you as well. >> >> >> >> >> -- >> >> --- >> 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 web2py+un...@googlegroups.com. >> For more options, visit https://groups.google.com/groups/opt_out. >> >> > > > -- > > --- > 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 web2py+unsubscr...@googlegroups.com. > For more options, visit https://groups.google.com/groups/opt_out. > > -- --- 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 web2py+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/groups/opt_out.
