FastCGI and ModPerl

classic Classic list List threaded Threaded
3 messages Options
Reply | Threaded
Open this post in threaded view
|

FastCGI and ModPerl

Scooter Willis-2
FastCGI and ModPerl
Now that I have everything setup in Gbrowse2 and working(almost) I am looking at the apache2 processes that are being spawned and looks like I have interesting behavior with ModPerl using the mgb2 option. Using the yeast_advance example when running fastcgi via fgb2 performance is quick and I don’t see any additional apache2 processes started. When loading the yeast_advance example using (mod_perl) mgb2 option every action appears to startup a collection of apache2 processes. The running apache2 process list gets very large very quickly and doesn’t appear to be normal. I also need to do a killall –9 apache2 to restart apache which is another indicator something isn’t working correctly.

In my particular gbrowse2 application I am loading everything into memory and when I checked this morning the system was using almost all of the allocated 8 gb of memory. Looking at the processes I had 50+ apache2 processes each allocating 400MB of memory. The behavior appears to be the same as the yeast_advance example but with a much higher penalty because of the amount of memory each apache2 process consumes. I have zero knowledge about perl and even less knowledge about ModPerl. Any help or guidance would be appreciated.

I tried using fastcgi option for the genome I configured but get an Internal Server Error and nothing descriptive in the log files. Any tricks or config settings required to get fastcgi working? It works for yeast_advance example but not for my particular genome.

Thanks

Scooter

------------------------------------------------------------------------------
This SF.net email is sponsored by Sprint
What will you do first with EVO, the first 4G phone?
Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
_______________________________________________
Gmod-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gmod-devel
Reply | Threaded
Open this post in threaded view
|

Re: FastCGI and ModPerl

Scott Cain
Hi Scooter,

I'm cc'ing this to the gbrowse mailing list, so more eyes can get a look at it.

Scott


On Wed, Jun 30, 2010 at 10:12 AM, Scooter Willis <[hidden email]> wrote:

>
> Now that I have everything setup in Gbrowse2 and working(almost) I am
> looking at the apache2 processes that are being spawned and looks like I
> have interesting behavior with ModPerl using the mgb2 option. Using the
> yeast_advance example when running fastcgi via fgb2 performance is quick and
> I don’t see any additional apache2 processes started. When loading the
> yeast_advance example using (mod_perl) mgb2 option every action appears to
> startup a collection of apache2 processes. The running apache2 process list
> gets very large very quickly and doesn’t appear to be normal. I also need to
> do a killall –9 apache2 to restart apache which is another indicator
> something isn’t working correctly.
>
> In my particular gbrowse2 application I am loading everything into memory
> and when I checked this morning the system was using almost all of the
> allocated 8 gb of memory. Looking at the processes I had 50+ apache2
> processes each allocating 400MB of memory. The behavior appears to be the
> same as the yeast_advance example but with a much higher penalty because of
> the amount of memory each apache2 process consumes. I have zero knowledge
> about perl and even less knowledge about ModPerl. Any help or guidance would
> be appreciated.
>
> I tried using fastcgi option for the genome I configured but get an Internal
> Server Error and nothing descriptive in the log files. Any tricks or config
> settings required to get fastcgi working? It works for yeast_advance example
> but not for my particular genome.
>
> Thanks
>
> Scooter
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by Sprint
> What will you do first with EVO, the first 4G phone?
> Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
> _______________________________________________
> Gmod-devel mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/gmod-devel
>
>



--
------------------------------------------------------------------------
Scott Cain, Ph. D.                                   scott at scottcain dot net
GMOD Coordinator (http://gmod.org/)                     216-392-3087
Ontario Institute for Cancer Research

------------------------------------------------------------------------------
This SF.net email is sponsored by Sprint
What will you do first with EVO, the first 4G phone?
Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
_______________________________________________
Gmod-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gmod-devel
Reply | Threaded
Open this post in threaded view
|

Re: FastCGI and ModPerl

Siddhartha Basu
In reply to this post by Scooter Willis-2
Hi Scooter,

On Wed, 30 Jun 2010, Scooter Willis wrote:

>    Now that I have everything setup in Gbrowse2 and working(almost) I am
>    looking at the apache2 processes that are being spawned and looks like I
>    have interesting behavior with ModPerl using the mgb2 option. Using the
>    yeast_advance example when running fastcgi via fgb2 performance is quick
>    and I don't see any additional apache2 processes started. When loading the
>    yeast_advance example using (mod_perl) mgb2 option every action appears to
>    startup a collection of apache2 processes. The running apache2 process
>    list gets very large very quickly and doesn't appear to be normal. I also
>    need to do a killall -9 apache2 to restart apache which is another
>    indicator something isn't working correctly.

You might have to look at your apache2 configuration. I am not about
your OS and setup but here are couple of things that might influence the
behaviour ....

* What MPM of apache2 being used ?
* MaxClients and MaxRequestsPerChild at least if you use worker MPM ?
* Are these apache2 process shows up as soon as the server starts or
  under concurrent requests.
* I would suggest test with *ab* with various no of concurrent requests
 and see which one pushes it to the limit. Then tune up those apache2
 params and keep on testing with *ab* to get to a reasonable setup that
 you are satisfied with.

Hope this helps.
-siddhartha

>
>    In my particular gbrowse2 application I am loading everything into memory
>    and when I checked this morning the system was using almost all of the
>    allocated 8 gb of memory. Looking at the processes I had 50+ apache2
>    processes each allocating 400MB of memory. The behavior appears to be the
>    same as the yeast_advance example but with a much higher penalty because
>    of the amount of memory each apache2 process consumes. I have zero
>    knowledge about perl and even less knowledge about ModPerl. Any help or
>    guidance would be appreciated.
>
>    I tried using fastcgi option for the genome I configured but get an
>    Internal Server Error and nothing descriptive in the log files. Any tricks
>    or config settings required to get fastcgi working? It works for
>    yeast_advance example but not for my particular genome.
>
>    Thanks
>
>    Scooter

> ------------------------------------------------------------------------------
> This SF.net email is sponsored by Sprint
> What will you do first with EVO, the first 4G phone?
> Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
> _______________________________________________
> Gmod-devel mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/gmod-devel


------------------------------------------------------------------------------
This SF.net email is sponsored by Sprint
What will you do first with EVO, the first 4G phone?
Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
_______________________________________________
Gmod-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gmod-devel