[Catalyst] Development environments and performance

John Goulah jgoulah at gmail.com
Wed Jan 16 17:13:06 GMT 2008


On Jan 16, 2008 5:05 AM, Ian Docherty <catalyst at iandocherty.com> wrote:

> I would just like to canvas opinion on a good development environment
> and practices for Catalyst since we are starting to have some
> performance issues.
>
> We are running on a managed server (Red Hat Enterprise), Apache2, MySql
> 5.1
>
> We have several people developing on this machine (about 5), each with
> their own sandboxes.
>
> I have set up apache so that each developer runs their Catalyst
> application under fast-cgi.
>
> We are beginning to see some performance issues especially with several
> people testing at the same time on different fast-cgi processes.
>
> The 'top' command shows 'server.pl' and 'perl' taking the bulk of the
> cpu time.
>
> Best case for a simple dynamic page in Catalyst to respond is about 5
> seconds. With several people testing at the same time it can take  30
> seconds or more to respond to a simple page. I would have hoped for only
> a second or so on such a lightly loaded server.
>
> so.
> Is the problem likely to be the setup we are using? (we can't use
> mod-perl since I believe we can't run multiple versions of the same
> application)



Why wouldn't you just use the standalone server bundled with Catalyst?  Fcgi
is great for production, but the processes are fairly thick memory wise, so
having instances for each developer could be an issue.   We use the cat
server for development and works fine for about 5-10 people at any given
time on a modest box (4G ram)




>
>
> Could we have problems in the performance of our application. Should we
> start thinking about profiling it to see where the time is going? (If
> so, are there any pointers on how best to do this?)



As mentioned, the catalyst server gives you  a lot of good numbers to
measure performance.


Good luck,
John
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.scsys.co.uk/pipermail/catalyst/attachments/20080116/458bf=
eea/attachment.htm


More information about the Catalyst mailing list