Re: [Trac] Trac (almost) hangs

2010-11-12 Thread Torsten Bronger
Hallöchen!

Torsten Bronger writes:

 Our trac installation on Ubuntu 10.04 server is extremely slow.
 Every request (no matter whether POST or GET,
 ticket/roadmap/search/etc) takes approx. 200 seconds.

I did some packet sniffing, and LDAP communication is the problem.
I switched authentication off and Trac ran smoothly.  It's weird
because a Django installation running on the same Apache uses LDAP
too and has no problems.

Anyway, it doesn't seem to have anything to do with Trac.

Tschö,
Torsten.

-- 
Torsten Bronger, aquisgrana, europa vetus
   Jabber ID: torsten.bron...@jabber.rwth-aachen.de
  or http://bronger-jmp.appspot.com

-- 
You received this message because you are subscribed to the Google Groups Trac 
Users group.
To post to this group, send email to trac-us...@googlegroups.com.
To unsubscribe from this group, send email to 
trac-users+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/trac-users?hl=en.



Re: [Trac] Trac (almost) hangs

2010-11-10 Thread Dave Love
Matthew Caron matt.ca...@sixnet.com
writes:

 On 11/09/2010 06:05 AM, Dave Love wrote:
 I guess it's https://bugs.launchpad.net/trac/+bug/607117

 I'm still on 8.04, so this very well could be it. I haven't tried it on yet.

On a quiet system, run trac-admin under strace to see what's going on
(assuming that normally pauses, as it did for me).

-- 
You received this message because you are subscribed to the Google Groups Trac 
Users group.
To post to this group, send email to trac-us...@googlegroups.com.
To unsubscribe from this group, send email to 
trac-users+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/trac-users?hl=en.



Re: [Trac] Trac (almost) hangs

2010-11-09 Thread Dave Love
Torsten Bronger writes:

 Hallöchen!

 Our trac installation on Ubuntu 10.04 server is extremely slow.
 Every request (no matter whether POST or GET,
 ticket/roadmap/search/etc) takes approx. 200 seconds.

 After the first very slow response, trac becomes fast, but if you
 stop browsing for a while, the next request is slow again.  (As if
 it must be waked up.)

I guess it's https://bugs.launchpad.net/trac/+bug/607117

-- 
You received this message because you are subscribed to the Google Groups Trac 
Users group.
To post to this group, send email to trac-us...@googlegroups.com.
To unsubscribe from this group, send email to 
trac-users+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/trac-users?hl=en.



Re: [Trac] Trac (almost) hangs

2010-11-09 Thread Matthew Caron

On 11/09/2010 06:05 AM, Dave Love wrote:

I guess it's https://bugs.launchpad.net/trac/+bug/607117


I'm still on 8.04, so this very well could be it. I haven't tried it on yet.

--
Matthew Caron
Build Engineer
Sixnet | www.sixnet.com
O +1 518 877 5173 Ext. 138
F +1 518 602 9209
matt.ca...@sixnet.com

--
You received this message because you are subscribed to the Google Groups Trac 
Users group.
To post to this group, send email to trac-us...@googlegroups.com.
To unsubscribe from this group, send email to 
trac-users+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/trac-users?hl=en.



[Trac] Trac (almost) hangs

2010-11-08 Thread Torsten Bronger
Hallöchen!

Our trac installation on Ubuntu 10.04 server is extremely slow.
Every request (no matter whether POST or GET,
ticket/roadmap/search/etc) takes approx. 200 seconds.

After the first very slow response, trac becomes fast, but if you
stop browsing for a while, the next request is slow again.  (As if
it must be waked up.)

CPU is not used while waiting for trac.  The harddisk load is
negligible, too.

Any idea why this is?


We use the following configuration:
http://wilson.homeunix.com/about.html

Apache's httpd.conf says:

--8---cut here---start-8---
WSGIScriptAlias /trac /home/chantal/trac.wsgi
WSGIScriptAlias / /home/chantal/django.wsgi
ServerName chantal.ipv.kfa-juelich.de:80

MinSpareServers 2
MaxSpareServers 5
StartServers 1
MaxClients 40

KeepAlive Off
#KeepAliveTimeout 2
MaxKeepAliveRequests 200

XSendFile on
XSendFileAllowAbove on

AllowEncodedSlashes On

Location /trac
  AuthType Basic
  AuthBasicProvider ldap
  AuthzLDAPAuthoritative off
  AuthUserFile /dev/null
  AuthLDAPUrl 
ldaps://dc-e02.ad.fz-juelich.de:636/DC=ad,DC=fz-juelich,DC=de?sAMAccountName?sub?(department=IEK-5)
  AuthName Authorization required
  require valid-user
/Location
--8---cut here---end---8---

Tschö,
Torsten.

-- 
Torsten Bronger, aquisgrana, europa vetus
   Jabber ID: torsten.bron...@jabber.rwth-aachen.de
  or http://bronger-jmp.appspot.com

-- 
You received this message because you are subscribed to the Google Groups Trac 
Users group.
To post to this group, send email to trac-us...@googlegroups.com.
To unsubscribe from this group, send email to 
trac-users+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/trac-users?hl=en.



Re: [Trac] Trac (almost) hangs

2010-11-08 Thread Matthew Caron

On 11/08/2010 03:48 AM, Torsten Bronger wrote:

Any idea why this is?


I believe it is because you aren't starting enough servers/keeping 
enough spare servers


My out of the box Ubuntu 10.04 server setup is:

IfModule mpm_prefork_module
StartServers  5
MinSpareServers   5
MaxSpareServers  10
MaxClients  150
MaxRequestsPerChild   0
/IfModule

IfModule mpm_worker_module
StartServers  2
MaxClients  150
MinSpareThreads  25
MaxSpareThreads  75
ThreadsPerChild  25
MaxRequestsPerChild   0
/IfModule

Which section you use depends, of course, in which version of apache 
you're running (the traditional prefork model, or the new threaded 
worker model).


Other thoughts:

- If the machine is RAM constrained, it is possible that any files read 
into memory and cached there are being dropped as the machine sits idle, 
so it needs to re-read them from memory.


- If the browser's cache times out and the page content is flushed, all 
the graphics, etc. will need to be pulled in again. Setting up a static 
cache of these so they are pulled through apache rather than through 
wsgi/trac/etc. may help things


- If you're using SQLite as a DB, and have several users, you will 
likely see issues like this if a user is reading from a table which is 
being updated by another user.

--
Matthew Caron
Build Engineer
Sixnet | www.sixnet.com
O +1 518 877 5173 Ext. 138
F +1 518 602 9209
matt.ca...@sixnet.com

--
You received this message because you are subscribed to the Google Groups Trac 
Users group.
To post to this group, send email to trac-us...@googlegroups.com.
To unsubscribe from this group, send email to 
trac-users+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/trac-users?hl=en.



Re: [Trac] Trac (almost) hangs

2010-11-08 Thread Torsten Bronger
Hallöchen!

Matthew Caron writes:

 On 11/08/2010 03:48 AM, Torsten Bronger wrote:

 Any idea why this is?

 I believe it is because you aren't starting enough servers/keeping
 enough spare servers

I don't believe that it could take 200 seconds (did you perhaps read
milliseconds?) to launch another apache process.  For me, it seems
that some code waits for a timeout.

I switched on trac's logging but there are no messages.

Tschö,
Torsten.

-- 
Torsten Bronger, aquisgrana, europa vetus
   Jabber ID: torsten.bron...@jabber.rwth-aachen.de
  or http://bronger-jmp.appspot.com

-- 
You received this message because you are subscribed to the Google Groups Trac 
Users group.
To post to this group, send email to trac-us...@googlegroups.com.
To unsubscribe from this group, send email to 
trac-users+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/trac-users?hl=en.