[Tilecache] Preponderance of "stuck lock" errors when running tilecache_seed.py

Christopher Schmidt crschmidt at metacarta.com
Thu Jan 8 14:55:44 EST 2009


On Thu, Jan 08, 2009 at 11:31:25AM -0800, Roger André wrote:
> Ok, so then if MetaTiling is used, the tilecache cannot be viewed while it
> is being seeded. Correct?

Typically, the TileCache 'lock' sticks around for 30 secnds before it
thinks it is 'stuck'. This is configurable in the cache section of the
cnofig file, to be longer. 

If your 'individual' metatiles take mre than 30 seconds to render, you
should bump this up.

If your 'stuck lock' messages are occurring sooner than 30 seconds, then
you probably have a bug. Patches welcome, in that case.

-- Chris

> 
> On Thu, Jan 8, 2009 at 11:23 AM, Christopher Schmidt <
> crschmidt at metacarta.com> wrote:
> 
> > On Thu, Jan 08, 2009 at 09:44:12AM -0800, Roger André wrote:
> > > I'm curious why I'm seeing so many of these errors now?  I'm using
> > Tilecache
> > > against a MapServerLayer to seed my cache, and am getting these errors
> > > whenever I try to view the web page at the same time as the seeding is
> > > taking place.  This did not seem to be a problem when the layer was a WMS
> > > one, and I wasn't using metatiling.
> >
> > MetaTiling is the *only* thing that uses locks. Without meta-tiling,
> > locks would never happen.
> >
> > Regards,
> > --
> > Christopher Schmidt
> > MetaCarta
> >

-- 
Christopher Schmidt
MetaCarta



More information about the Tilecache mailing list