[rrd-users] rrdtool 1.3.0 - memory eaten away by mmap.

[Tobias Oetiker]

Hi Sebastian,

Today Sebastian Harl wrote:

> Hi Tobi,
>
> On Wed, Jun 18, 2008 at 04:50:30PM +0200, Tobias Oetiker wrote:
> > Today Sebastian Harl wrote:
> > > On Wed, Jun 18, 2008 at 03:13:59PM +0200, Tobias Oetiker wrote:
> > > > Yesterday Raimund Berger wrote:
> > > > > So I sent the rrdtool process SIGABRT and dumped core when those
> > > > > mmap's happen. Apparently, it turns out to be some issue with
> > > > > libcairo. Even if it's input is faulty, it sure shouldn't behave like
> > > > > that.
> > > > >
> > > > > Anyway, back trace attached.
> > > >
> > > > glad to hear it is cairo, so that question is what makes it go belly
> > > > up ... maybe we can alter rrdtool NOT to stroke it in that
> > > > particular way (and I thought by moving away from libart I got rid
> > > > of these problems ... )
> > >
> > > I've once had a problem with libcairo (from Debian Etch iirc) that would
> > > segfault if I did too much drawing (some tens of thousands of updates
> > > iirc) without calling cr_stroke() in between. I cannot image that
> > > rrdtool would draw that much but I thought I'd mention it anyway ;-)
> > the drawing code in rrdtool draws each graph LINE,AREA as a single
> > large shape. So depending on the size of the image this could
> > potentially have a lot of points ... not 10k but still more than
> > just a few.
>
> Okay - we could double check if that's the same issue by adding
> cr_stroke() after, say, 100 drawing commands. I'm not entirely sure if
> it was in the 10k's in my case, it might as well have been (large)
> 1k's...
>
> As Raimund cannot reproduce it on Debian Sid as well, we've got another
> indication that those two issues seem to be related.

If it is the strokeing, then the problem should go away when you
call the same graph with AREA instead of LINE

cheers
tobi

>
> Cheers,
> Sebastian
>
>

-- 
Tobi Oetiker, OETIKER+PARTNER AG, Aarweg 15 CH-4600 Olten
http://it.oetiker.ch tobi at oetiker.ch ++41 62 213 9902