Hey Pieter On 9/15/06, Pieter Rautenbach <[EMAIL PROTECTED]> wrote: > We're trying to use PyMSNt in a production environment with ejabberd, but are > experiencing a lot of problems. The transport will run for a (seemingly) > random > period of time before it will simply stop responding. The python process > refuses > to respond to a normal kill signal, meaning that one has to force it to stop > with kill -9. > ... > It looks like PyMSNt went into a loop. > > Has anyone experienced similar behaviour? Any help or hints will be > appreciated. > I should add that this is my first contact with PyMSNt.
I've been running my yahoo and irc transports under WinPdb [1] to help figure out where they hang up and die. It's "An Advanced Python Debugger" that has a backend/frontend split that allows you to run the python code and the debugger on difference machines, with different os'es. If you can organise it, I strongly suggest you try this to figure out where the tight loop is being created. I know it sucks to have to 'debug' a live transport with thousands of users (hey I'm one of them!), but some times it's the only way. The _rpdb2 process only seems to add a 5% processing overhead, so it's well worth it. (I think it does have quite a large memory overhead, so you might want to watch that one) [1] http://www.digitalpeers.com/pythondebugger/ -- - Norman Rasmussen - Email: [EMAIL PROTECTED] - Home page: http://norman.rasmussen.co.za/ From [EMAIL PROTECTED] Fri Sep 15 11:28:44 2006 From: [EMAIL PROTECTED] (Pieter Rautenbach) Date: Fri Sep 15 11:26:15 2006 Subject: [py-transports] Self-induced flooding causes PyMSNt to stop responding Message-ID: <[EMAIL PROTECTED]> ******************************************************************* Click here to view our e-mail legal notice: http://www.mxit.co.za/pdfs/mxit_legal.pdf or call: +27 21 888 7920 ******************************************************************* Hi Norman, Thank you for the idea -- I'm having a look at it now. In the mean time, any other comments on this issue are welcome! Pieter -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Norman Rasmussen Sent: 15 September 2006 12:41 PM To: PyTransports Discussion Subject: Re: [py-transports] Self-induced flooding causes PyMSNt to stopresponding Hey Pieter On 9/15/06, Pieter Rautenbach <[EMAIL PROTECTED]> wrote: > We're trying to use PyMSNt in a production environment with ejabberd, but are > experiencing a lot of problems. The transport will run for a (seemingly) random > period of time before it will simply stop responding. The python process refuses > to respond to a normal kill signal, meaning that one has to force it to stop > with kill -9. > ... > It looks like PyMSNt went into a loop. > > Has anyone experienced similar behaviour? Any help or hints will be appreciated. > I should add that this is my first contact with PyMSNt. I've been running my yahoo and irc transports under WinPdb [1] to help figure out where they hang up and die. It's "An Advanced Python Debugger" that has a backend/frontend split that allows you to run the python code and the debugger on difference machines, with different os'es. If you can organise it, I strongly suggest you try this to figure out where the tight loop is being created. I know it sucks to have to 'debug' a live transport with thousands of users (hey I'm one of them!), but some times it's the only way. The _rpdb2 process only seems to add a 5% processing overhead, so it's well worth it. (I think it does have quite a large memory overhead, so you might want to watch that one) [1] http://www.digitalpeers.com/pythondebugger/ -- - Norman Rasmussen - Email: [EMAIL PROTECTED] - Home page: http://norman.rasmussen.co.za/ _______________________________________________ py-transports mailing list [email protected] http://www.modevia.com/cgi-bin/mailman/listinfo/py-transports
