Subject: Marine Geospatial Ecology Tools (MGET) help
Text archives
| From: | "Jason Roberts" <> |
|---|---|
| To: | "'Trish Miller'" <> |
| Cc: | <> |
| Subject: | RE: [mget-help] error message |
| Date: | Tue, 15 Jun 2010 10:32:07 -0400 |
|
Hi Trish, I’m glad it’s working now. In theory, the code that is reading the rasters should be accessing
them in a way that allows more than one program to read from them at the same
time, but nobody to write to them. But I did not write that code and have never
looked at it, so I cannot be 100% sure. Perhaps the code was written to be more
conservative and open the rasters with “exclusive” access,
forbidding any other programs to read from or write to them. But I doubt that
would be done intentionally in this case. A more likely possibility is that there was a problem accessing
the server. This could happen if there was a temporary network glitch (likely
on wireless, much less likely with a wired network) or if the server reached
its connection limit or was overloaded. These latter two cases are unlikely if
it is a real server running a server operating system. If it is running Windows
XP, Vista, or 7 though, then it is possible to encounter a problem with three
computers reading from it. I think the connection limit with those operating
systems is 5, but you might be able to hit it with less than 5 computers if any
of them are opening multiple connections to the machine. Anyway, thanks for the positive feedback and good luck. Jason From: Trish Miller
[mailto:] Hi Jason- On Fri, Jun 11, 2010 at 11:29 AM, Jason Roberts <> wrote: Hi Trish, We’re sorry you’re seeing that
error. It comes from the R rgdal package, which is being used to read the
prediction rasters into memory. Obviously the message is not very helpful so
we’ll have to do some detective work to figure out what is happening. Early versions of the tool could fail by
running out of memory because they tried to read the entire raster into memory
at once. This usually resulted in a different error and that problem has been
fixed since MGET 0.6 or so. Most likely it is not a memory issue. Some suggestions to try: 1. Verify that ArcGIS itself can read the raster by previewing it
with ArcCatalog. Most likely you have already done this. 2. Reduce the extent of the prediction region by providing a smaller
template raster. The offset number is very large, suggesting your prediction raster
is very large. The tool should work on any size raster providing that rgdal can
read it, but maybe you are hitting a limit and rgdal is not reporting a clear
message. 3. Enable verbose logging, reproduce the problem, and email me the
entire output from the geoprocessing window and I’ll see if I can spot
anything. There are instructions for enabling verbose logging here: http://code.env.duke.edu/projects/mget/wiki/Configuring%20MGET%20logging Best regards, Jason From: Trish
Miller [mailto:]
Hello- |