Search results matching tag 'RubyOnRails'

Re: Limitations regarding InstanceMaxRequests and IdleTimeout IIS6 fcgiext 1.5

bg — Sun, 18 Oct 2009 10:23:05 GMT

Dear WenJun Zhang,

Thank you for answer. It was indeed the application pool´s idle shutdown feature that caused the shut down of the idle processes.

I have now created a new application pool for my RoR site, trying to find settings that give acceptable performance for my application. Avoiding idle timeout of the Ruby processes in my pool helped to some degree but I have so far not been able to prevent the processes from dying after about 120 requests, regardless of the load I put on the IIS. So this is still a major problem.

The typical behaviour when testing the site with low load is a response time, when a ruby process in the pool starts, of 10 seconds. The following consecutive requests render a response in 90-120 ms, until the process dies and starts over. One Ruby process occupies about 35 Mb of memory, and there is plenty of memory on my server to accommodate many Ruby processes at the same time. During my resent tests I have limited the number of instances to 4. (MaxInstances=4)

In my application pool I now run with these settings:

Recycling panel:
Recycle worker processes (in minutes): unchecked,
Recycle worker processes (number of requests): unchecked,
Recycle worker processes at the following times: unchecked,
Memory recycling: Maximum virtual memory: unchecked,
Maximum used memory: unchecked

Performance panel:
Idle timeout: unchecked,
Request queue limit: unchecked,
Enable CPU monitoring: unchecked, maximum number of worker processes: 4

Health panel:
Enable pinging: checked,
Enable rapid-fail protection, unchecked,
Startup time limit: 100,
Shutdown time limit: 90

Fcgiext.init:
ExePath=D:\ruby\bin\ruby.exe
Arguments=D:\rails\Ruby\public\dispatch.fcgi
IgnoreDirectories=0
IgnoreExistingFiles=1
QueueLength=1000
MaxInstances=4
InstanceMaxRequests=100000
IdleTimeout=43200
StderrMode=IgnoreAndReturn200
ResponseBufferLimit=8388608
EnvironmentVars=RAILS_ENV:production

Are there any other settings I could try to manipulate in order to make my ruby processes survive more than 120 requests, even at low load conditions (two requests per second)

I really would appreciate suggestions how to resolve this problem.

Thanks.

Simultaneous requests to Rails

chrisortman — Wed, 05 Sep 2007 17:04:40 GMT

I've managed to get a rails app working under IIS using FastCGI, however I've run into a bit of a snag. As a test I have a client program send 10 simultaneous requests to the same rails page (simple render 'hello world')

However when I slowly ramp up and get 10 ruby processes sitting idle on the webserver I can get all 10 simultaneous requests to complete successfully. I'm not sure if I would want a way to always have at least N workers or if there needs to be a way to get ruby.exe to start faster?

Thanks

Error 0x80004005 occurred processing request.

dfunky — Tue, 04 Sep 2007 22:26:00 GMT

I followed "10 steps to get Ruby on Rails running on Windows with IIS FastCGI" to get my app running on two different W2K3 servers.

Everything went fine with the first installation but I just can't get it work on the second server. I've repeated the steps over and over and I still get the same error:

Error 0x80004005 occurred processing request.

The FastCGI process exited unexpectedly

This is what I get in the fastcgi.crash.log after trying the test controller:

[04/Sep/2007:18:13:51 :: 6296] Ignoring unsupported signal USR2.
[04/Sep/2007:18:13:51 :: 6296] Ignoring unsupported signal HUP.
[04/Sep/2007:18:13:51 :: 6296] Ignoring unsupported signal SIGTRAP.
[04/Sep/2007:18:13:51 :: 6296] starting
[04/Sep/2007:18:13:51 :: 6296] Dispatcher failed to catch: You have a nil object when you didn't expect it!
You might have expected an instance of Array.
The error occurred while evaluating nil.split (NoMethodError)
C:/Ruby/lib/ruby/1.8/cgi.rb:897:in `parse'
C:/Ruby/lib/ruby/gems/1.8/gems/actionpack-1.13.3/lib/action_controller/cgi_ext/raw_post_data_fix.rb:45:in `initialize_query'
C:/Ruby/lib/ruby/1.8/cgi.rb:2274:in `initialize'
C:/Ruby/lib/ruby/site_ruby/1.8/fcgi.rb:592:in `new'
C:/Ruby/lib/ruby/site_ruby/1.8/fcgi.rb:592:in `each_cgi'
C:/Ruby/lib/ruby/gems/1.8/gems/rails-1.2.3/lib/fcgi_handler.rb:141:in `process_each_request!'
C:/Ruby/lib/ruby/gems/1.8/gems/rails-1.2.3/lib/fcgi_handler.rb:55:in `process!'
C:/Ruby/lib/ruby/gems/1.8/gems/rails-1.2.3/lib/fcgi_handler.rb:25:in `process!'
E:/aweb/LocalUser/xxxx/public/dispatch.fcgi:24
almost killed by this error
[04/Sep/2007:18:13:51 :: 6296] Dispatcher failed to catch: You have a nil object when you didn't expect it!
You might have expected an instance of Array.
The error occurred while evaluating nil.split (NoMethodError)
C:/Ruby/lib/ruby/1.8/cgi.rb:897:in `parse'
C:/Ruby/lib/ruby/gems/1.8/gems/actionpack-1.13.3/lib/action_controller/cgi_ext/raw_post_data_fix.rb:45:in `initialize_query'
C:/Ruby/lib/ruby/1.8/cgi.rb:2274:in `initialize'
C:/Ruby/lib/ruby/site_ruby/1.8/fcgi.rb:592:in `new'
C:/Ruby/lib/ruby/site_ruby/1.8/fcgi.rb:592:in `each_cgi'
C:/Ruby/lib/ruby/gems/1.8/gems/rails-1.2.3/lib/fcgi_handler.rb:141:in `process_each_request!'
C:/Ruby/lib/ruby/gems/1.8/gems/rails-1.2.3/lib/fcgi_handler.rb:55:in `process!'
C:/Ruby/lib/ruby/gems/1.8/gems/rails-1.2.3/lib/fcgi_handler.rb:25:in `process!'
E:/aweb/LocalUser/xxxx/public/dispatch.fcgi:24
killed by this error

The only thing that went different from the first installation to the other was the "MSVCR71.dll was not found" error I got durring the 2nd install. Still the "require 'fcgi'" instruction resulted to true so I assumed to installation was ok...

Please help me !

FastCGI with Ruby on Rails - IIS 5.1 / IIS 6 - Http POST issue.

mspeight — Fri, 15 Jun 2007 01:30:51 GMT

I have got IIS and Multiple App's to talk using same technique’s as I used with isap_fcgi.dll, all works fine apart from an intermittent problem.

After returning data from a POST, the rubyw.exe process is left in some funny state, and the next request causes rubyw.exe to exit.

The same request may work 3 times, then crash on the 4^th.

Steps are as follows POST -> Returns a Redirect -> GET on Redirfect fails to return (ruby.exe exits) -> Browser has another go new instance of ruby.exe started and page returned

Other Steps are POST -> Returns just render_text -> Another browser instance goes to url -> No page/error is returned (i.e. the new request got to rubyw but it crashed)

I have duplicated this with a simple form, that posts and returns itself.

It seems to be a timing thing, But I cant figure out the cause.

PS: I have patched the cgi.rb (To remove the IIS reference, if you don’t you get random crashes ay any time)

Anyone come across this before.

-----------------

By set-up is detailed here http://mspeight.blogspot.com/2007/05/how-to-run-multiple-rails-apps-under.html (but uses fcgiext.dll instead of isapi_fcgi.dll). I did need to repatch the "AbstractRequest:request_uri" differently as this FastCGI does provide REQUEST_URI

Directory Listing Denied

armanddp — Mon, 28 May 2007 22:26:16 GMT

I know this is probably not directly related to the FastCGI handler but after configuring my site in IIS to support Ruby on Rails following the 10 steps instructions over here I can't get the default document to work. i.e., when people browse to http://mysite.com/ they get directory listing denied but when browsing to a controller fastcgi kicks in and my Rail application works.

I know I'm missing something obvious but for the life of me can't get it to serve a default document. I've tried various default document combinations. Tried the isapirewrite4 filter to rewrite requests to dispatch.fcgi and in a short ritual offered a small chicken to the web gods but still no luck.

Any suggestions would really be appreciated. (Great work on the fastcgi handler, apart from the default doc issue it seems to be working like a charm)

Cannot disable Ruby logging when using fastcgi?

Stoo — Sun, 01 Apr 2007 20:20:40 GMT

I'm not sure if the problem I have is something that can be fixed on the fastcgi end of things, but thought I would post to see if anyone else had come across the same issue...

When I set the log_level to :none in my rails app's config, It will not start when running through fastcgi. The rails app will still start from command line Ruby, and will also start using the Mongrel web service gem for Ruby. It only fails when calling it through FastCGI.

If I set the logging to anything other than ":none" (like :info), then everything works just fine.

Sadly, under windows, Ruby only writes one log file, and never releases the lock on that file until you kill the app... Makes disk space a wee bit of an issue when the site needs to be up 24/7 :-)

Any pointers in the right direction greatly appreciated as always. - especially as this may not be something that can be fixed in the fastcgi component!

Setting a PATH per site with FastCGI

mollo — Thu, 15 Mar 2007 14:29:31 GMT

Hi,

Thats a very good new for running Ror Apps under IIS.

I need to run multiple Rails apps under IIS with TP2, theses apps require different Ruby version (because of some compatibilty issues with ImageMagick, Oracle and so..)

How specify a PATH variable per IIS site handled by TP2.

One app need c:\ruby\bin

Other need c:\ruby185\bin

Thanks.

Getting Ruby to stay in memory under fastcgi.

Stoo — Tue, 13 Mar 2007 22:27:52 GMT

First up, let me just say that this FastCGI for IIS has solved many stability issues that we had with older fastcgi products.

We do have one last issue to resolve however. We run Ruby on Rails under FastCGI on IIS6.0 (2k3SP1). We also run Postgres on these servers, but nothing else.

When we make our first request to the server, it takes a few seconds for the Rubyw.exe to start up and compile our code. Once it is started, everything appears to be fairly snappy and generally runs well. Our problem is that after around 5 minutes of being idle, the rubyw.exe process terminates. Another one will start as soon as a request is made, but this makes for a fairly slow browsing experience for our users as they often have to wait for the ruby app to "spin up".

Is there a way to keep these ruby processes running in memory for a lot longer? I've tried tweaking the "InstanceTimeout" value in fcgiext.ini, but this appears to have no effect whatsoever, and the processes still seem to terminate after about 5 minutes.

Any help, or clues that may lead me in the right direction are muchly appreciated! We've solved so many other problems along the way that we've had, this seems to be the last little hurdle to using this solution in production!

Stoo

Where can I find error logs for the FCGI handler

mcmoyer — Mon, 05 Feb 2007 21:53:38 GMT

I'm currently trying to set up FastCGI with Ruby on rails. I haven't gotten very far :)

Anyway, I haven't progressed beyond this error message:

FastCGI Handler Extension

Error 0x80004005 occurred processing request.

The FastCGI process exited unexpectedly

I'm trying to find where the extension might be logging to. Thanks!

Marlon

Re: Support for other application frameworks

pboling — Fri, 08 Dec 2006 19:10:25 GMT

Mike,

Thanks for the response!

Before your response I had posted about this to the rails core team mailing list (I am not on the team, just on the list) and it has recieved some response, so you may want to check it out (it's a Google Groups mailing list):

http://groups.google.com/group/rubyonrails-core/browse_thread/thread/b88ad36ddd82477e

I would love to test, and help however I can. Is there a more recent version than the 'technical preview' I have already downloaded? I am a Mac / Linux guy, and am a bit shaky in Windows, but I am deploying my site on Windows Server 2003 so I need to learn it. Right now I am using mongrel as the webserver because it 'just works'. But I need to find a better solution because mongrel has no NTLM authentication module.

Thanks,

Peter