Jim Freels
mechanical side of nuclear engineering, multiphysics analysis, COMSOL specialist
Please login with a confirmed email address before reporting spam
Posted:
7 years ago
Dec 6, 2017, 4:51 p.m. EST
In linux, you stop a running process with the kill command. In windows, you can use the task manager or an equivalent command. Then you will start your batch job again using a command line similar to this:
comsol batch -inputfile fn.ext -outputfile fn2.ext -recover -continue
Use the comsol batch -help to see all the options available in the batch mode of comsol. Also, see the COMSOL documentation on batch.
Batch options:
-alivetime <seconds> The time between writing status on disc
-batchlog <log filename> File to store log in
-checklicense <filename> Print license requirement for an mph file
-client Run as client
-continue Continue computation
-dev <filename> Path to a JAR-file with additional classes
to call from the batch class file
-error <{on}|off> Stop if an error occurs
-graphics Display graphics
-host <hostname> Connect to host <hostname>
-inputfile <filename> The input file name (.mph or .class)
-job <job name> The batch job to run
-nosave Do not save the resulting model
-outputfile <filename> The output file name (if not specified the
input file name is used)
-paramfile <filename> Table file containing parameter names in
the first row and parameter values
in the following rows
-pindex <parameter indices> Comma separated list of parameter indices
-plist <parameter values> Comma separated list of parameter values
-pname <parameter name> Comma separated list of parameter names
-port <port number> Connect to port <port number>
-recover Recover and continue computation
-stoptime <seconds> The time the batch job is allowed to run
before it is stopped
-study <study name> The study to compute
-------------------
James D. Freels, Ph.D., P.E.
In linux, you stop a running process with the kill command. In windows, you can use the task manager or an equivalent command. Then you will start your batch job again using a command line similar to this:
comsol batch -inputfile fn.ext -outputfile fn2.ext -recover -continue
Use the comsol batch -help to see all the options available in the batch mode of comsol. Also, see the COMSOL documentation on batch.
Batch options:
-alivetime The time between writing status on disc
-batchlog File to store log in
-checklicense Print license requirement for an mph file
-client Run as client
-continue Continue computation
-dev Path to a JAR-file with additional classes
to call from the batch class file
-error Stop if an error occurs
-graphics Display graphics
-host Connect to host
-inputfile The input file name (.mph or .class)
-job The batch job to run
-nosave Do not save the resulting model
-outputfile The output file name (if not specified the
input file name is used)
-paramfile Table file containing parameter names in
the first row and parameter values
in the following rows
-pindex Comma separated list of parameter indices
-plist Comma separated list of parameter values
-pname Comma separated list of parameter names
-port Connect to port
-recover Recover and continue computation
-stoptime The time the batch job is allowed to run
before it is stopped
-study The study to compute
Please login with a confirmed email address before reporting spam
Posted:
7 years ago
Dec 7, 2017, 1:35 a.m. EST
Thank you,
I used "comsol -h" not "comsol batch -h", that was sloppy of me.
A follow up question is that if I have run an analysis and it has either finished, been terminate by me or not converged and I want to continue with a new study from the command line (perhaps with other parameters for time incrementation, different mass scaling, using a different solver, using different BC etc.). How do I do that?
[I want to do something similar as in Abaqus, where using *restart, read in the inp-file (call it newjob) and then abq -job=newjob -oldjob=terminated_job from the command line]
/Staffan
Thank you,
I used "comsol -h" not "comsol batch -h", that was sloppy of me.
A follow up question is that if I have run an analysis and it has either finished, been terminate by me or not converged and I want to continue with a new study from the command line (perhaps with other parameters for time incrementation, different mass scaling, using a different solver, using different BC etc.). How do I do that?
[I want to do something similar as in Abaqus, where using \*restart, read in the inp-file (call it newjob) and then abq -job=newjob -oldjob=terminated\_job from the command line]
/Staffan