McIDAS-V Support Forums

Hi -

Running the most recent nighty under Linux, McV does not read MSG HRIT files with a local server. It issues the following error message:
Encountered a problem (server: 'localhost:8112'):
Error reading times:
java.io.EOFException

Running the same nightly under Windows the files read just fine. Also, under Linux I can read local MODIS files, i.e. it does not appear that there is a generic problem with local ADDE servers.

cheers, HP

Hi HP,

Thanks for reporting this. I picked up some MSG HRIT files from the EUMETSAT EO Portal and I'm getting them to work fine on Windows 10 and macOS using today's nightly. I don't currently have access to a Linux machine that I can test on, so I forwarded this to a programmer with a Linux VM who will try things on his end. In the meantime, when you see the error, could you click the Support Form button in the Error window and submit a support request that way? This will give us some extra information/files that might help us determine what is going on.

Thanks,
Bob Carp

Hi HP,

Thanks for your support request. To this point, I haven't been able to replicate the error. Would you mind zipping up the directory containing your data and putting it on our ftp site for me to grab? Here are instructions:

ftp ftp.ssec.wisc.edu
user: anonymous
password: your email address
bin
cd pub/incoming
put filename
bye

Please let me know the name of the file(s) once they're posted and I'll give them a try on my end.

Thanks,
Bob

Hi HP,

We wanted to verify that you have the prologue/epilogue files in your data directory. Also, as an alternative to posting the files to ftp, you could give us a listing of the file names in the data directory for the local dataset. From the file names I should be able to order the same data from the EUMETSAT EO Portal myself and give them a try.

Thanks,
Bob

Hi Bob -
I have uploaded a 3-slot sample of 15-min SEVIRI compressed scans. File name is:
HRIT3slots5min.tar
cheers, HP

Hi HP,

After untarring the file, I see that the data is in native format (*.nat):

MSG4-SEVI-MSG15-0100-NA-20200723101242.999000000Z-20200723101300-1416378.nat
MSG4-SEVI-MSG15-0100-NA-20200723102743.047000000Z-20200723102800-1416378.nat
MSG4-SEVI-MSG15-0100-NA-20200723104243.096000000Z-20200723104301-1416378.nat

There are servers in McIDAS-X to handle native format data, but it turns out that those servers aren't yet available in McIDAS-V. I wrote this up as inquiry 2916. Until the native format server gets added, I'd suggest selecting the "HRIT data sets in tar file" format when placing your data in the EUMETSAT EO Portal. At this point, you should see that the MSG HRIT local server works in McIDAS-V.

Please let me know if you have any additional questions/concerns about this.

Thanks,
Bob

Sorry Bob, I have uploaded the wrong data set. Uploaded now
SEVIRI_15min.tar.gz
that contains 3 slots in HRIT format.
cheers, HP

Hi HP,

Thanks for the new files. Those are working for me as well as a local dataset. Would you mind starting a fresh run of McIDAS-V? From a terminal window, cd to your McIDAS-V-System directory and run:

runMcV -tempuserpath

This will start up McV as if it was the first time ever run on the machine. Could you try re-creating your local dataset and see if it works there? This would let us know if maybe some setting, plugin, or code in your Jython Library could somehow be interfering with things.

I doubt this would be the case since you have other local datasets working, but it could be worth verifying the permissions of the directory containing the data. I know that we've had cases in the past where users tried creating a local dataset where they didn't have full permissions to the data directory and they ran into problems.

Thanks again,
Bob

Hi Bob-

No way with starting native McV. I get the same error. I also checked/changed the folder permission without success.
My OS is Kubuntu 20.04 with KDE Plasma 5.18.5, 64-bit.
regards, HP

Hi HP,

Sorry for the delayed response. I just wanted to let you know that I'm still looking into this trying to figure out what's going on. I've still been unable to replicate the error you are seeing with your local dataset. Could you please verify that your data directory contains only the "H-000-MSG4*" files extracted from the zipped file you put on our ftp server? I found that if I create a dataset pointing to the zipped file I get your error.

Perhaps sending an "ls -l" output of your data directory could be useful.

Thanks again,
Bob