Materials Studio checks most of the data and settings required to perform a CASTEP job prior to launch. If it cannot start the job, error messages are generated detailing the reasons.
However, sometimes jobs may fail for reasons which cannot be checked prior to launch. In such cases, more detailed information about the error
may only be available in the
[seedname].castep file produced by the job or in the project log file. Other files stored in
the job directory on the server, for example,
Castepexe.log, may also contain further clues. To view the server-side files, you can use the
Remote View facility of the
Below is a list of the most common reasons for CASTEP jobs to fail. It may help you to identify and fix any problems you have with your remote CASTEP jobs. For generic reasons for remote job failures please consult the If a remote job fails help topic.
Select View | Project Log from the menu bar to see if any error or warning messages have been reported.
Common reasons for a CASTEP job to fail to start
Depending upon the options selected, CASTEP may use significant amounts of disk space to store scratch files. Scratch files are created by each
node during the execution of a parallel CASTEP job. CASTEP uses the value of the environment variable
GATEWAY_TMP as the location
to be used to save these files; this variable is set by
share/bin/ms_vars.sbd and can be changed using the gateway's
web interface. You should ensure that the location that
will be used on each node points to a file system with at least 1 GB of free space. Note that the
./tmp setting for
GATEWAY_TMP corresponds to using the common file space on the head node, in the actual job directory, to store the temporary files.
This setting can have a detrimental effect on the performance of Linux clusters.
An additional problem may occur if the
GATEWAY_TMP is set to the location on a shared disk resource which is NFS mounted from
the nodes. This mount should be done in a synchronous mode using hard mounts, as detailed in
Installing Materials Studio on Linux systems.
It is important to identify the reason for the failure of a CASTEP job before taking any action.
In most cases, the error message produced will give a good indication of the reason for the failure. If the error message indicates that the job
failed, but does not provide specific reasons, check the
[seedname].castep file produced by the job or check the project log.