Rails - Feb 2006 - Performance profiling & routes

Hi all,

With a rails app running under lighttpd, we''ve recently stopped using
lighttpd''s URL rewriting in favour of letting Rails'' routes do
their job.

However, under proper testing, we now see the application slowing down after
a period of time (a couple of days); pages take increasingly longer to load
over time.  This is fixed by a lighttpd restart.

Could this performance degredation have anything to do with the use of the
routes (we''re just using the default set)?  If not, can anyone offer
any
pointers about good ways to track down performance issues in a Rails app?

Please let me know if you need any more info.

Thanks much for your help,
 - Tim
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
http://wrath.rubyonrails.org/pipermail/rails/attachments/20060209/8410928f/attachment.html

How is the machine''s health when the application slows down?

I don''t have extensive experience with production Rails apps, but
there seems to be a clear pattern of ever increasing memory usage
causing machines to being swapping.

There are some good emails in the archive about how to force
Ruby garbage collection. Perhaps you should incorporate forced
collection into your application...

--  
-- Tom Mornini


On Feb 8, 2006, at 7:08 PM, Tim Riley wrote:
> Hi all,
>
> With a rails app running under lighttpd, we''ve recently stopped  
> using lighttpd''s URL rewriting in favour of letting
Rails'' routes
> do their job.
>
> However, under proper testing, we now see the application slowing  
> down after a period of time (a couple of days); pages take  
> increasingly longer to load over time.  This is fixed by a lighttpd  
> restart.
>
> Could this performance degredation have anything to do with the use  
> of the routes (we''re just using the default set)?  If not, can  
> anyone offer any pointers about good ways to track down performance  
> issues in a Rails app?
>
> Please let me know if you need any more info.
>
> Thanks much for your help,
>  - Tim
> _______________________________________________
> Rails mailing list
> Rails@lists.rubyonrails.org
> http://lists.rubyonrails.org/mailman/listinfo/rails

Also make certain that you are running in production and not  
development mode(i''m sure you did this already). Also what kind of  
rails app is it? Does it use RMagick at all? Other extensions?

	I don''t think that routes are causing the problem especially if you  
are just using the defaults. I have rails apps running on lighttpd/ 
fcgi for months without reboots that are stable in their memory  
consumption. But I have other apps that use RMagick heavily that  
would steadily increase in memory usage.

	With a little more info I might be able to help more? Does the app  
just get slower or does the memory usage grow? Also have you  
benchmarked the app after a restart and then when it gets slow?  
Sometime perceptions can be wrong. Also, what amount of traffic does  
the site get? Rails apps that run on fcgi and end up stiing idle for  
hours ata time can be slow to start again because the fcgi processes  
get paged out of memory. If this is the case I have solved it by  
using wget and a cron job to hit the site every 5 minutes to keep the  
fcgi''s warmed up.

Cheers-
-Ezra


On Feb 8, 2006, at 10:13 PM, Tom Mornini wrote:
> How is the machine''s health when the application slows down?
>
> I don''t have extensive experience with production Rails apps, but
> there seems to be a clear pattern of ever increasing memory usage
> causing machines to being swapping.
>
> There are some good emails in the archive about how to force
> Ruby garbage collection. Perhaps you should incorporate forced
> collection into your application...
>
> -- 
> -- Tom Mornini
>
>
> On Feb 8, 2006, at 7:08 PM, Tim Riley wrote:
>
>> Hi all,
>>
>> With a rails app running under lighttpd, we''ve recently
stopped
>> using lighttpd''s URL rewriting in favour of letting
Rails'' routes
>> do their job.
>>
>> However, under proper testing, we now see the application slowing  
>> down after a period of time (a couple of days); pages take  
>> increasingly longer to load over time.  This is fixed by a  
>> lighttpd restart.
>>
>> Could this performance degredation have anything to do with the  
>> use of the routes (we''re just using the default set)?  If not,
can
>> anyone offer any pointers about good ways to track down  
>> performance issues in a Rails app?
>>
>> Please let me know if you need any more info.
>>
>> Thanks much for your help,
>>  - Tim
>> _______________________________________________
>> Rails mailing list
>> Rails@lists.rubyonrails.org
>> http://lists.rubyonrails.org/mailman/listinfo/rails
>
> _______________________________________________
> Rails mailing list
> Rails@lists.rubyonrails.org
> http://lists.rubyonrails.org/mailman/listinfo/rails
>
-Ezra Zygmuntowicz
Yakima Herald-Republic
WebMaster
http://yakimaherald.com
509-577-7732
ezra@yakima-herald.com