How does apache runs a application when a request comes in?

Question

I have a python web application running on apache2 deployed with mod_wsgi. The application has a thread continuously running. This thread is a ZeroMQ thread and listening to a port in loop. The application is not maintaining session. Now if I open the browser and sends a request to the apache server the data is accepted for the first time. Now when second time I send the request It shows Internal server error. When I checked the error log file for traceback, It shows the ZMQError:- The address already in use.

Does apache reloads the application on each request sent from the browser since so that the ZeroMQ thread is being created everytime and being assigned the port but since the port has already been assigned it shows error....

Answer 1

It looks like your application is using zmq to bind to some port.

As you have suspected already, each request can be run as independent process, thus competing in access to the port to bind to.

There can be so called workers, each running one process processing http/wsgi requests, and each trying to bind.

You shall redesign your app not to use bind, but connect, this will probably require having another process with zeromq serving something you do with that (but this last line is dependent on what you do in your app).

Comments

You mean to say that I should put the ZMQ logic in some another process and each worker process of the apache2 when handling the request should talk to it?? Right?

Answer 2

It sure sounds like that's what's happening. Have you tried putting some logging statements in your code so that you can see what's going on?

If someone else here knows more about mod_wsgi he can update us. Failing that, it seems that mod_wsgi is specifically supported over at:

https://groups.google.com/forum/#!forum/modwsgi

Comments

Yes I Have put logging information in the code and everything is working fine except this part of it.

Answer 3

Long story short, yes. But it's more complicated. Apache can do one of several things to handle multiple requests, by pre-forking subprocesses of by spinning up threads within a listening process or a combination of both. If you first process is still running when your second request comes in, it will fail as you've seen. However, even if it terminates the kernel will hold that socket in a FINWAIT state for some amount of time, usually 2 minutes in Linux.

You might try to keep the process running indefinitely, but poorly written web application can leak resources and quickly cause httpd to crash. Because of this httpd is very strict about enforcing execution
time limits on processes invoked to handle web responses and forcibly killing long running threads. It's actually quite difficult to get a long running process to spawn off from httpd.

I did this for one project, I fork/exec'd off a long running process, then returned a link to the user that when clicked would communicate with the forked process via an RPCXML connection it would listen on to
return the status of the job. It was extremely difficult to get apache to not kill the process, you essentially had to follow all the steps of creating a daemon process to disassociate the subprocess from the apache
daemon, I then coded a watchdog thread to be sure it killed itself off after some amount of time.

A better design would probably be to start you ZMQ process independently of the server, then use something like XMLRPC or something RESTful to allow web requests to communicate with it and quickly receive the results they need. Not knowing more about your requirements, that's the best I can propose.

Hope this helps.

Comments

Thanks, I got what u are saying. It's really helpful.

---

I got a solution by using this
WsgiDaemonProcess processes=1 threads=10
Now do u think it will work
Also Not requests are accepted at the same time so i think it should work. And in fact at present it is working fine.