2.3.3 Check Queue status
Using the Clusterware Pro V5.1 Web Interface:
From the Jobs tab:
z
Review the Queues table. Use the Previous and Next buttons to view more
Queues.
Using the Clusterware Pro V5.1 CLI:
% bqueues [<queue_name>]
References:
z
3.7.8 How do I access the Clusterware Pro V5.1 Web Interface?
z
3.7.9 How do I access the Clusterware Pro V5.1 Command Line Interface?
Back to Top
2.3.4 Remove temporary files from Compute Nodes
The process of removing temporary files created by an application can vary depending on
whether the job was submitted to an AppRS queue or a non-AppRS LSF queue.
Jobs submitted to an AppRS queue execute in temporary directories allocated by AppRS on
each node where the job runs. AppRS copies the contents of these directories back to the host
that the job was submitted from when the job completes. As long as the application only
generates files within its execution directory, there is no need for the user to remove
temporary files generated by an application.
In the event AppRS restarts an application on a new set of nodes, the original working
directories and files created before the migration are not removed. This is done in order to be
as careful as possible about avoiding data loss. In addition, the files and directories on the
unavailable host(s) cannot be removed at that time because the host cannot be reached. The
user must manually remove these files using the AppRS commands apprs_ls and
apprs_clean. Use apprs_ls to list the directories allocated by AppRS to the current user:
%
apprs_ls
For non-root users, apprs_clean can only be used to remove files that belong to a specific job
id. This protects files that are currently in use by another job from being accidentally erased.
%
apprs_clean <jobid>
root can delete all apprs related directories. This command should be used with great caution
To Next Page
Loading...
Need help?
General