[gdal-dev] RFC 26: GDAL Block Cache Improvements

[Tamas Szekeres]

Devs,

I'm quite involved in fixing the problems have already been raised in
http://n2.nabble.com/Problems-with-large-raster-sizes-WMS-TMS-td3996139.html

Those issues are mostly related to the following issues:

1. The current GDAL block cache implementation may result in out of memory
errors with large raster dimensions.
2. The raster size is stored in a signed int in GDAL which may overflow when
the size of the raster is large.


These issues will less likely happen with most drivers storing the raster in
the same physical file. However this may be a problem with those drivers
which compose the rasters from multiple tiles which eventually results in
rasters with large virtual dimensions.

I'm about to handle issue #1 by introducing RFC26 which would establish a
hastable based caching solution in parallel to the current array based block
cache implementation.
As this change would affect the gdal core it seems reasonable to come up
with a new RFC here and open up a discussion about that:

http://trac.osgeo.org/gdal/wiki/rfc26_blockcache

You could also review the corresponding patch which implements this
solution:

http://trac.osgeo.org/gdal/attachment/ticket/3264/blockcache.patch#preview

Any comments or ideas would be helpful.

Best regards,

Tamas
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.osgeo.org/pipermail/gdal-dev/attachments/20091206/3647dc1e/attachment.html