Long-Running Processes
We understand Long-Running Processes to be remote jobs that run without interactive konsole use and over long periods of time. Typically scientific simulations or other complex computations.
Please notify us if you intend to run such jobs.
We offer these options for this class of job:
- Concurrent use of our Computer Pools
- Exclusive use and reservation of one or more Pool Computers
- Dedicated Virtual Machine(s)
If your compute demands cannot be met with these options, we recommend the PC² as an alternative. You can find a general overview on the PC² here:
- https://pc2.uni-paderborn.de/hpc-services/available-systems/available-hpc-systems
- https://uni-paderborn.atlassian.net/wiki/spaces/PC2DOK/overview
Further information on how to apply for cluster resources and how to access these cluster resources is available here:
- https://pc2.uni-paderborn.de/hpc-services/our-services/system-access-application
- https://uni-paderborn.atlassian.net/wiki/spaces/PC2DOK/pages/12944205/Access+and+Login
Using Pool Computers for Long Running Jobs
We generally allow concurrent use of our Pool Computers for smaller long runners as long as lokal interaktive users, and especially CS Department use, is not impaired.
Please take note of the following, though:
- You should inform us if you intend to run jobs for several consecutive hours.
- We offer absolutely no guarantee that your jobs complete uninterruptedly.
- We reserve the right to terminate all jobs without prior notice.
- The IRB Service page offers remote Wake-On-Lan power on of our computers.
- Your Kerberos Ticket expires 10 hours after your login. You may renew it for up to 7 days. See kinit(1) for details.