Hi Even,<br><br> Thanks for the pointer - might try again with the -multi option. Anyway, the -co TILED=YES is speeding things up drastically (something like a factor 10).<br><br> Cheers,<br> Simon<br><br><br><div class="gmail_quote">
On Mon, Feb 25, 2013 at 1:16 PM, Even Rouault <span dir="ltr"><<a href="mailto:even.rouault@mines-paris.org" target="_blank">even.rouault@mines-paris.org</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
Le lundi 25 février 2013 11:13:47, Simon Lyngby Kokkendorff a écrit :<br>
<div class="im">> Hi Even,<br>
><br>
> > When you say "crashed", you mean it exited with the integer overflow<br>
> > error (to<br>
> > be opposed as the windows that is displayed by Windows when a process<br>
> > really<br>
> > crashes) ?<br>
><br>
> Yes, it didn't really crash, but exited with the integer overflow error<br>
> message.<br>
><br>
> The issue is rather with the -wm 3072. The warping algorithm will currently<br>
><br>
> > cast a integer overflow error when a memory allocation above 2 GB is<br>
> > attempted.<br>
> > So you should try -wm 2047 or less. There's rarely a significant<br>
> > advantage in<br>
> > using so big values for -wm.<br>
><br>
> Ahh, I see. I was using win64 and just assumed that setting -vm high might<br>
> speed things up.<br>
<br>
</div>See<br>
<a href="http://trac.osgeo.org/gdal/wiki/UserDocs/GdalWarp#WillincreasingRAMincreasethespeedofgdalwarp" target="_blank">http://trac.osgeo.org/gdal/wiki/UserDocs/GdalWarp#WillincreasingRAMincreasethespeedofgdalwarp</a><br>
<div class="im"><br>
> Will try to create a tiled output image :-)<br>
> By the way, I also tried using -multi to speed up (with otherwise the same<br>
> options, I think), but got an error saying something like "Unable to aquire<br>
> IOMutex" (unfortunately I haven't got the exact output).<br>
<br>
</div>Hum, this might be a side effect of a particular high value of -wm which cause<br>
particularly big, and consequently long, I/O operations. The mutex timeouts at<br>
10 minutes currently. Usually unitary I/O operations only last a few seconds<br>
or dozains seconds.<br>
<br>
><br>
> Thanks,<br>
> Simon<br>
</blockquote></div><br>