ePN and memory leaks

[Ethan Galstad]

I think they memory leaks may actually be occurring in the embedded 
interpreter itself, although I can't say for sure.  Nagios 2.0 will 
have an option to reload the interpreter after a predefined number of 
calls.  Hopefully this will "fix" the memory leak problem, even if it 
is a bit of a hackish thing to do.


On 14 Jan 2003 at 10:24, Stanley Hopcroft wrote:

> Dear Sir,
> 
> I am writing to thank you for your helpful letter and say,
> 
> On Mon, Jan 13, 2003 at 03:32:00PM -0500, atonns at mail.ivillage.com wrote:
> > Has anyone else done an in-depth investigation of memory leaks in ePN and/or
> > custom checks written for ePN?
> 
> > > -----Original Message-----
> > > From: Stanley Hopcroft [mailto:Stanley.Hopcroft at IPAustralia.Gov.AU]
> > > Sent: Monday, December 30, 2002 10:31 PM
> > > To: atonns at mail.ivillage.com
> > > Cc: nagios-users at lists.sourceforge.net
> > > Subject: Re: [Nagios-users] ePN and memory leaks
> > > 
> > > > Are there any suggestions for finding memory leaks in ePN? 
> > > (I'm going to be
> > > > looking at Apache::Leak and Devel::Leak, but there might be 
> > > other debug
> > > > options in nagios I'm not aware of).
> > > >
> > > 
> > > Not that I am aware of. It sounds like you are on the right track
> > > though.
> > 
> > Well, I've run several tests with both Apache::Leak and Devel::Leak and have
> > found that it's not my code that's leaking, but either p1.pl or mini_epn.
> > The number of SVs before and after executing my code stays the same (or
> > decreases!). BUT the number of SVs between invocations when called by
> > mini_epn increases. Below is an example where I use Devel::Leak and have
> > changed the output of my service check to be the ENTER and LEAVE values
> > returned by NoteSV and CheckSV (note that the first invocation increases
> > dramatically because of DynaLoaded modules)
> >
> 
> I am the author of mini_epn and am too aware (now that I have started 
> trying to code XS's) of how ignorantly it was coded.
> 
> Bear in mind that mini_epn is only a means of assessing whether or not a
> plugin will __run__ under an ePN (if it fails mini_epn, it will
> certainly fail ePN. If it passes however, it may still fail to run under
> ePN. There exist modules for example that will run once in mini_epn but 
> not more than once).
> 
> I will try and re-examine it and see if there are any faults that I can 
> see.
> 
> However, if there are faults in mini_epn, there are almost certainly 
> faults in ePN since mini_epn is the guts of ePN (from base/checks.c) 
> stuck in a loop.
> 
> Please fix them and it as best you may. 
> 
> The ePN stuff is _all_ contributed. The developer, Stephen Davies, (I
> will write you offline with his address) is acknowledged in the code (eg
> common/config.h.in) and in the docs.
> 
> Although Mr Galstad says he only includes contributions that he 
> understands and meet a need, this one may have slipped by.
>  
> > $ ./mini_epn 2>/dev/null
> > Enter file name: check_remote_interfaces-leak_test
> > embedded perl plugin output was -12820724,ENTER: 32488 LEAVE:32674
> > 
> > Enter file name: check_remote_interfaces-leak_test
> > embedded perl plugin output was -12820724,ENTER: 32686 LEAVE:32686
> > 
> > Enter file name: check_remote_interfaces-leak_test
> > embedded perl plugin output was -12820724,ENTER: 32694 LEAVE:32693
> > 
> > Enter file name: check_remote_interfaces-leak_test
> > embedded perl plugin output was -12820724,ENTER: 32701 LEAVE:32700
> > 
> > Enter file name: check_remote_interfaces-leak_test
> > embedded perl plugin output was -12820724,ENTER: 32708 LEAVE:32707
> > 
> > > You could also run any suggestions past the ePN author (noted in the
> > > source).
> > 
> > I couldn't find this person's info (is it not Ethan Galstad?).
> > I don't want to bother who are no longer involved in the project.
> > 
> 
> He can choose whether he wants to reply or not. As you are aobviously 
> interested and able to contribute to this stuff, I am sure he will 
> welcome it.
> 
> Yours sincerely.
> -- 
> ------------------------------------------------------------------------
> Stanley Hopcroft
> ------------------------------------------------------------------------
> 
> '...No man is an island, entire of itself; every man is a piece of the
> continent, a part of the main. If a clod be washed away by the sea,
> Europe is the less, as well as if a promontory were, as well as if a
> manor of thy friend's or of thine own were. Any man's death diminishes
> me, because I am involved in mankind; and therefore never send to know
> for whom the bell tolls; it tolls for thee...'
> 
> from Meditation 17, J Donne.
> 
> 
> -------------------------------------------------------
> This SF.NET email is sponsored by: FREE  SSL Guide from Thawte
> are you planning your Web Server Security? Click here to get a FREE
> Thawte SSL guide and find the answers to all your  SSL security issues.
> http://ads.sourceforge.net/cgi-bin/redirect.pl?thaw0026en
> _______________________________________________
> Nagios-users mailing list
> Nagios-users at lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/nagios-users
> 



Ethan Galstad,
Nagios Developer
---
Email: nagios at nagios.org
Website: http://www.nagios.org



-------------------------------------------------------
This SF.NET email is sponsored by: FREE  SSL Guide from Thawte
are you planning your Web Server Security? Click here to get a FREE
Thawte SSL guide and find the answers to all your  SSL security issues.
http://ads.sourceforge.net/cgi-bin/redirect.pl?thaw0026en