Hi Ryne, Thanks a lot a lot for the tip. David
On Thursday, July 21, 2016 at 6:10:58 PM UTC+2, Ryne Everett wrote: > > A couple causes of gunicorn timeouts I've seen is larger files that don't > transfer within the default --timeout (30 seconds). This isn't necessarily > your problem but if it is you can increase the timeout. I've also seen it > when streaming files, in which case the solution is to make the workers > asynchronous (-k gevent). Going from memory here so no promises. > > On Thu, Jul 21, 2016 at 12:08 PM, loevdav <[email protected] > <javascript:>> wrote: > >> Hi Ken, >> >> Thanks. >> >> Yes I can but I'm not sure where do you want to get because the issuet >> happens on different sites and on different actions (sometimes just loading >> the home page and other time isnide the admin). >> >> I have been looking on the web and it looks like some people solved it by >> increasing the number of workers for gunicorn. But these sites have very >> very low traffic and do not have any complex queries or operations. >> >> The more complex sites are on a server of their own and I have no issue. >> >> Anyway thanks and let me know what you want to try with runserver. >> >> Thanks >> David >> >> >> On Thursday, July 21, 2016 at 4:16:39 PM UTC+2, Kenneth Bolton wrote: >>> >>> Are you able to use the django runserver? It is very hard for me to tell >>> where the problem might be, so I would step through each piece until I find >>> the root cause, starting with the django runserver. >>> >>> hth, >>> >>> ken >>> >>> On Thu, Jul 21, 2016 at 6:06 AM, loevdav <[email protected]> wrote: >>> >>>> Hi Ken, >>>> >>>> Thanks a lot for the reply: >>>> >>>> Here are the logs: >>>> >>>> >>>> *Error log* >>>> >>>> [2016-07-01 10:30:43 +0000] [3496] [CRITICAL] WORKER TIMEOUT (pid:1105) >>>> [2016-07-01 10:30:49 +0000] [3496] [CRITICAL] WORKER TIMEOUT (pid:6783) >>>> [2016-07-01 10:35:53 +0000] [3496] [CRITICAL] WORKER TIMEOUT (pid:7129) >>>> [2016-07-01 10:41:32 +0000] [3496] [CRITICAL] WORKER TIMEOUT (pid:7317) >>>> [2016-07-01 11:25:37 +0000] [3496] [CRITICAL] WORKER TIMEOUT (pid:7802) >>>> [2016-07-01 11:30:44 +0000] [3496] [CRITICAL] WORKER TIMEOUT (pid:8969) >>>> [2016-07-01 12:45:47 +0000] [3496] [CRITICAL] WORKER TIMEOUT (pid:11499) >>>> [2016-07-01 15:25:46 +0000] [3496] [CRITICAL] WORKER TIMEOUT (pid:17183) >>>> [2016-07-01 15:30:43 +0000] [3496] [CRITICAL] WORKER TIMEOUT (pid:17363) >>>> [2016-07-06 23:00:35 +0000] [3496] [CRITICAL] WORKER TIMEOUT (pid:27959) >>>> [2016-07-08 09:25:56 +0000] [3496] [CRITICAL] WORKER TIMEOUT (pid:13078) >>>> [2016-07-08 09:30:50 +0000] [3496] [CRITICAL] WORKER TIMEOUT (pid:13287) >>>> [2016-07-08 09:40:54 +0000] [3496] [CRITICAL] WORKER TIMEOUT (pid:13675) >>>> [2016-07-08 09:45:52 +0000] [3496] [CRITICAL] WORKER TIMEOUT (pid:13876) >>>> [2016-07-08 10:45:47 +0000] [3496] [CRITICAL] WORKER TIMEOUT (pid:14081) >>>> [2016-07-08 10:55:52 +0000] [3496] [CRITICAL] WORKER TIMEOUT (pid:16225) >>>> [2016-07-13 11:05:51 +0000] [3496] [CRITICAL] WORKER TIMEOUT (pid:23348) >>>> [2016-07-13 11:15:53 +0000] [3496] [CRITICAL] WORKER TIMEOUT (pid:24537) >>>> [2016-07-13 11:20:47 +0000] [3496] [CRITICAL] WORKER TIMEOUT (pid:24786) >>>> [2016-07-13 12:00:47 +0000] [3496] [CRITICAL] WORKER TIMEOUT (pid:28216) >>>> [2016-07-15 12:30:44 +0000] [3496] [CRITICAL] WORKER TIMEOUT (pid:29462) >>>> [2016-07-20 12:05:40 +0000] [3496] [CRITICAL] WORKER TIMEOUT (pid:19347) >>>> [2016-07-20 12:35:37 +0000] [3496] [CRITICAL] WORKER TIMEOUT (pid:25708) >>>> [2016-07-20 12:55:42 +0000] [3496] [CRITICAL] WORKER TIMEOUT (pid:30319) >>>> [2016-07-21 09:10:45 +0000] [3496] [CRITICAL] WORKER TIMEOUT (pid:15845) >>>> >>>> >>>> *nginx error log* >>>> >>>> 2016/04/27 21:17:00 [info] 30913#0: *266 client 79.44.98.211 closed >>>> keepalive connection >>>> 2016/04/27 21:17:00 [info] 30913#0: *268 client 79.44.98.211 closed >>>> keepalive connection >>>> 2016/04/27 21:17:00 [info] 30912#0: *270 client 79.44.98.211 closed >>>> keepalive connection >>>> 2016/04/27 21:17:00 [info] 30912#0: *272 client 79.44.98.211 closed >>>> keepalive connection >>>> 2016/04/27 21:17:00 [info] 30912#0: *274 client 79.44.98.211 closed >>>> keepalive connection >>>> 2016/04/27 21:17:00 [info] 30912#0: *276 client 79.44.98.211 closed >>>> keepalive connection >>>> 2016/04/27 21:17:00 [info] 30912#0: *278 client 79.44.98.211 closed >>>> keepalive connection >>>> 2016/04/27 21:17:00 [info] 30912#0: *280 client 79.44.98.211 closed >>>> keepalive connection >>>> 2016/04/27 21:17:00 [info] 30912#0: *282 client 79.44.98.211 closed >>>> keepalive connection >>>> 2016/04/27 21:17:42 [info] 30912#0: *295 client 79.44.98.211 closed >>>> keepalive connection >>>> 2016/04/27 21:17:42 [info] 30912#0: *301 client 79.44.98.211 closed >>>> keepalive connection >>>> >>>> *Supervisor log* >>>> >>>> /home/mars/.virtualenvs/domus/local/lib/python2.7/site-packages/mezzanine/pages/templatetags/pages_tags.py:134: >>>> >>>> RemovedInDjango110Warning: render() must be called with a dict, not a >>>> Context. >>>> return t.render(Context(context)) >>>> >>>> /home/mars/.virtualenvs/domus/local/lib/python2.7/site-packages/django/template/defaulttags.py:499: >>>> >>>> RemovedInDjango110Warning: Reversing by dotted path is deprecated >>>> (mezzanine.pages.views.page). >>>> url = reverse(view_name, args=args, kwargs=kwargs, >>>> current_app=current_app) >>>> >>>> /home/mars/.virtualenvs/domus/local/lib/python2.7/site-packages/mezzanine/template/__init__.py:122: >>>> >>>> RemovedInDjango110Warning: render() must be called with a dict, not a >>>> Context. >>>> return self.template.render(context) >>>> >>>> /home/mars/.virtualenvs/domus/local/lib/python2.7/site-packages/mezzanine/pages/templatetags/pages_tags.py:134: >>>> >>>> RemovedInDjango110Warning: render() must be called with a dict, not a >>>> Context. >>>> return t.render(Context(context)) >>>> >>>> /home/mars/.virtualenvs/domus/local/lib/python2.7/site-packages/django/template/defaulttags.py:499: >>>> >>>> RemovedInDjango110Warning: Reversing by dotted path is deprecated >>>> (mezzanine.pages.views.page). >>>> url = reverse(view_name, args=args, kwargs=kwargs, >>>> current_app=current_app) >>>> >>>> /home/mars/.virtualenvs/domus/local/lib/python2.7/site-packages/mezzanine/template/__init__.py:122: >>>> >>>> RemovedInDjango110Warning: render() must be called with a dict, not a >>>> Context. >>>> return self.template.render(context) >>>> >>>> /home/mars/.virtualenvs/domus/local/lib/python2.7/site-packages/mezzanine/pages/templatetags/pages_tags.py:134: >>>> >>>> RemovedInDjango110Warning: render() must be called with a dict, not a >>>> Context. >>>> return t.render(Context(context)) >>>> >>>> /home/mars/.virtualenvs/domus/local/lib/python2.7/site-packages/django/template/defaulttags.py:499: >>>> >>>> RemovedInDjango110Warning: Reversing by dotted path is deprecated >>>> (mezzanine.pages.views.page). >>>> url = reverse(view_name, args=args, kwargs=kwargs, >>>> current_app=current_app) >>>> >>>> >>>> On this server I have 10 low traffic website and 4GB RAM. >>>> >>>> Here is the RAM usage: >>>> >>>> total used free shared >>>> buffers cached >>>> Mem: 3998 3736 262 52 30 >>>> 150 >>>> -/+ buffers/cache: 3555 443 >>>> Swap: 255 255 0 >>>> >>>> >>>> Thanks a lot >>>> >>>> David >>>> >>>> On Wednesday, July 20, 2016 at 4:05:43 PM UTC+2, Kenneth Bolton wrote: >>>>> >>>>> Hi David, >>>>> >>>>> >>>>> I've gotten away with Mezzanine from the deploy scripts on some very >>>>> low-power hardware, including raspberry pi and virtualboxen with 500MB >>>>> RAM, >>>>> My guess is this is not an issue of resources. >>>>> >>>>> Can you share the relevant Django logs that coincide with your bad >>>>> gateway? They may provide insight. >>>>> >>>>> -ken >>>>> >>>>> On Wed, Jul 20, 2016 at 9:42 AM, loevdav <[email protected]> >>>>> wrote: >>>>> >>>>>> Hi, >>>>>> >>>>>> I got a lot of bad gateway on my sites build with mezzanine. >>>>>> >>>>>> They are hosted on a virtual server. >>>>>> >>>>>> >>>>>> I have added more ram but still got the issue (but with less >>>>>> occurrences). >>>>>> >>>>>> >>>>>> Previously I was using Apache (without issues) to deploy sites but >>>>>> lately I have been using Mezzanine deploy scripts. >>>>>> >>>>>> >>>>>> Does anyone knows about the minimum resources needed to run >>>>>> nginx/gunicorm per site? >>>>>> >>>>>> >>>>>> >>>>>> Thanks a lot >>>>>> >>>>>> -- >>>>>> You received this message because you are subscribed to the Google >>>>>> Groups "Mezzanine 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. >>>>>> >>>>> >>>>> -- >>>> You received this message because you are subscribed to the Google >>>> Groups "Mezzanine 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. >>>> >>> >>> -- >> You received this message because you are subscribed to the Google Groups >> "Mezzanine Users" group. >> To unsubscribe from this group and stop receiving emails from it, send an >> email to [email protected] <javascript:>. >> For more options, visit https://groups.google.com/d/optout. >> > > -- You received this message because you are subscribed to the Google Groups "Mezzanine 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.
