Re: Pool Module Question (Now Proxy Server Question)

When I run the proxy server directly from a console session, it eventually dies with a "Segmentation Fault" error message.  So the issue isn't with the daemonize scripts, but with my proxy server or spyce setup.  It happens under load, so it's probably just a poorly handled application condition that is bubbling up and taking down the proxy server.

On the birghter side, I've managed to work through my problems with Mod-Python, and have returned to that route.  The application had previously run under mod_python for two years, so I'm guessing this will work as a permanent solution.  

The performance seems to have improved with the upgrade to spyce 2.0.3, switching to a SQLAlchemy managed db pool, and taking advantage of the single-instance pickle pool.

Thanks again for all of the help Jonathan and Conan.

Also, is there a way to make a financial donation to Spyce?  

Ben Ringold
Hypothetical Software


---------- Original Message ----------------------------------
From: "Jonathan Ellis" <[email protected]>
Date:  Mon, 17 Apr 2006 09:26:22 -0700

>On Mon, 17 Apr 2006 09:53:26 -0600, "Ben Ringold"
><ben-3GbDDtM/[email protected]> said:
>> I've used the scripts Conan suggested with a few path changes (see
>> attached modified files).  The performance seems great, considerably
>> faster than the previous installation I had (spyce 1.3, mod_python with
>> poorly pooled db connections).  However, the proxy server dies off and
>> the daemon disappears every two hours or so.
>> 
>> I don't see anything useful in usiecr2.err, and usiecr2.log just has a
>> few generic exception messages before the daemon disappears:
>> 
>> Exception happened during processing of request from ('XXX.XXX.XXX.XXX',
>> YYYY)
>> 
>> I'm not sure if there are other logs that may have pertinent information,
>> or perhaps there are changes I could make to those daemonization scripts
>> to make it more robust.
>
>Ick, that's not cool.  I take it .err and .log are stderr and stdout
>redirections from the daemon script?
>
>I can't guess what's causing it to die, but I can suggest ways to try to
>narrow it down some more.
>
>- check the apache request log -- is it a specific page that kills it
>reproducibly?
>- use ab or a similar load-generating tool to just hammer it and see if
>it dies faster that way (remember to have ab use multiple "clients" --
>-c option?)
>- try running Spyce from a console instead of with the daemon script and
>see if it either crashes less or gives a useful exception
>
>-J
 




________________________________________________________________
Sent via the WebMail system at hypotheticalsoftware.com


 
                   


-------------------------------------------------------
This SF.Net email is sponsored by xPML, a groundbreaking scripting language
that extends applications into web and mobile media. Attend the live webcast
and join the prime developer group breaking into this new coding territory!
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=110944&bid=241720&dat=121642