Hi,
I am trying to do some jitter studies using Elegant. I would like to submit an array of jobs to run them in parallel. The result I get is that one of the output files (.cen or .twi, or any other) is locked for writing. This is an example of the error message I get:
Error:
unable to open file Aramis.cen for writing--file is locked (SDDS_InitializeOutput)
If I remove the output from these files I have the same kind of error in the coordinate and parameter outputs.
Maybe there is a setting I have to change to have the simulations run? Is there a setting of the cluster have I to modify (we recently migrated to a new one)?.
I uploaded the files.
Many thanks in advance for any feed-back.
Kind regards.
Run array of kobs
Moderators: cyao, michael_borland
-
- Posts: 45
- Joined: 06 Jul 2011, 03:14
Run array of kobs
- Attachments
-
- Elegant_array_job.7z
- (34.42 MiB) Downloaded 174 times
-
- Posts: 1951
- Joined: 19 May 2008, 09:33
- Location: Argonne National Laboratory
- Contact:
Re: Run array of kobs
Simona,
The problem is that your jobs all use the same destination for the output data. Each job needs to have different filenames. You can do this using the macro-substitution feature. E.g., in your .ele file, use
then on the commandline use
where "submit" is a script that submits a command to the batch queue. You'll end up with output files called Aramis_1.*, Aramis_2.*, ..., Aramis_10.*
--Michael
The problem is that your jobs all use the same destination for the output data. Each job needs to have different filenames. You can do this using the macro-substitution feature. E.g., in your .ele file, use
Code: Select all
rootname = Aramis_<index>
Code: Select all
for index in $(seq 1 10) ; do
submit elegant Aramis_ok.ele -macro=index=$index
done
--Michael
-
- Posts: 45
- Joined: 06 Jul 2011, 03:14
Re: Run array of kobs
Hi,
thanks for your answer. I tried but I was generating files named NAME_FILE_<INDEX>, where index was really the word "index". I asked to the expert of the cluster I am using, and he changed the script I use to submit the job:
CMD="elegant"
ARGS="Aramis_ok.ele "$@
in
CMD="elegant"
ARGS="Aramis_ok.ele "
In this case Elegant generates different files with the names NAME_FILE_1_*, NAME_FILE_2_*, NAME_FILE_3_*, ...
I tried to plot some files, and they look the same.
Many thanks for your help as usual.
thanks for your answer. I tried but I was generating files named NAME_FILE_<INDEX>, where index was really the word "index". I asked to the expert of the cluster I am using, and he changed the script I use to submit the job:
CMD="elegant"
ARGS="Aramis_ok.ele "$@
in
CMD="elegant"
ARGS="Aramis_ok.ele "
In this case Elegant generates different files with the names NAME_FILE_1_*, NAME_FILE_2_*, NAME_FILE_3_*, ...
I tried to plot some files, and they look the same.
Many thanks for your help as usual.