PACE maintenance is scheduled to start at 6:00 am Wednesday, February 3, 2020, and continue through Friday, February 5, 2020, at 11:59 pm. All headnodes and clusters will be unavailable while we make upgrades to improve overall PACE experience. More details of the activities this maintenance day will be available at the PACE blog at http://blog.pace.gatech.edu closer to the maintenance.
- We aim to provide as many resources as possible to you get started. Be sure to check out getting started for general information about logging into PACE managed clusters. We also have a frequently asked questions section for more general information. If you would like to contact us directly, there are two ways of reaching us:
Email us directly at firstname.lastname@example.org¶
- The more information we have, the faster we can resolve the problem. The following are very helpful:
- The username you use when logging in to the cluster (what you use to log into the Georgia Tech email)
- The exact error message that is displayed to the screen
- The exact sequence of commands you executed that resulted in that error
- The location of the job script or program you were trying to run
- An explanation of what you have already attempted
- Any symptoms that may be related to the problem
- If you are requesting software, please provide the URL to the software as well as the version you would like installed.
Submit a support request by using the support request script¶
- You can run the following script to request support:
pace-support.sh [jobid] [attachment1] [attachment2]
The jobid and attachments are optional. You can attach as many files as you wish, with or without specifying the jobid.
- gathers information commonly requested by administrators
- asks for you to type a message to be sent with the data
- attaches the given file(s) to the support request, if any (optional)
- emails email@example.com and your known GATech email address
Example Scenario: You submitted job "12345" with job script "my_job.pbs", but it isn't running when you think it should be. The very same job ran successfully in the past and output file "my_job.output." You could run the following:
pace-support.sh 12345 my_job.pbs my_job.output