[GRASS-user] /usr/lib/grass/scripts/r.in.wms: line 439: /usr/lib/grass/etc/r.in.wms/r.in.gdalwarp: Argument list too long

[Markus Neteler]

On Wed, May 27, 2009 at 12:49 AM, Paulo 'Pmarc' Marcondes
<pmarc.debian at gmail.com> wrote:
> Hi All,
>
> I was starting a project that needed a lot of srtms.
>
> So I though: Let's use the WMS!
>
> there I went, set region:
>
> GRASS 6.2.3 (prominence):~/gis > g.region -p
> projection: 3 (Latitude-Longitude)
> zone:       0
> datum:      wgs84
> ellipsoid:  wgs84
> north:      13N
> south:      56S
> west:       82W
> east:       34W
> nsres:      0:00:03
> ewres:      0:00:03
> rows:       82800
> cols:       57600
> cells:      4769280000
>
> then get the files:
> GRASS 6.2.3 (prominence):~/gis > r.in.wms output=srtm.jpl
> mapserver=http://wms.jpl.nasa.gov/wms.cgi layers=huemapped_srtm
>
> so, 4673 GTIFF files later, I got this error:
> /usr/lib/grass/scripts/r.in.wms: line 439:
> /usr/lib/grass/etc/r.in.wms/r.in.gdalwarp: Argument list too long
>
> now the qiestions:
> 1. what happened?
> 2. how to overcome this?
>
> I have thought to manually join all 4673 files, but then there has to
> be a (much) better way. I am open to suggestions.

This problem came up last year( in a different context):

On Mon, Oct 6, 2008 at 11:38 AM, Glynn Clements
<glynn at gclements.plus.com> wrote:
>> On Wed, Sep 24, 2008 at 5:04 PM, Daniel Victoria <daniel.victoria at gmail.com> wrote:
>> > I've bumped into this problem before, with the same error, but outside
>> > GRASS. Which shows that this is not a GRASS bug but a limitation of
>> > the operating system.
>>
> It's more robust to use sysconf(_SC_ARG_MAX), as the limit isn't
> necessarily fixed at compile time (e.g. it can vary between kernel
> versions).
>
> Also, the limit is divided between argv[] and the environment. If you
> have a lot of space taken up by environment variables, it will eat
> into the space available. This can be significant on older systems,
> where the limit may be as low as 4KiB (a single page of memory).
>

Possible solution:
- use r.in.wms from GRASS 7 (written in Python)
- amend the script to use shorter filenames
- use v.mkgrid and loop over the tiles to download in
   smaller chunks

Markus