Skip to content

Changes

Page history
post power outage authored by Doak, Peter W.'s avatar Doak, Peter W.
Show whitespace changes
Inline Side-by-side
Cades.md
View page @ 9d5c3ca1
......@@ -19,8 +19,13 @@ We *own* **1216** cores.
# Policies #
Walltime Limit: 48 hours
Simultaneous Jobs: 6
### Walltime Limit: 48 hours
### Simultaneous Jobs: 5
### Max processors * remaining seconds running at anytime: 27648000 or 640 cores for 12 hours.
If you need this relaxed to do your science [email](mailto:doakpw@ornl.gov).
If this is only because you need to run large numbers of identical or very long calculations consider whether NERSC or OLCF would be a better place to do "production."
* Run test jobs to establish timing and parallelization parameters.
* Try to make your walltimes tight. (not always 48:00:00).
......@@ -64,9 +69,11 @@ There is only one non-experimental queue: batch
You want to use the qos: **std**
Unless you have a short development job i.e. a test or debug run, then qos:**devel**
See this gitlab repo's examples for the pbs commands, there are more than on most clusters and they matter.
If you need to run wide relatively short jobs, are experiencing long waits for std and can deal with them being occassionally prempted (i.e. killed) you can request access to qos: **burst** via [XCAMS](https://xcams.ornl.gov/xcams/groups/cades-cnms-burst)
This is the obligatory PBS header for a job.
### Basic PBS header
``` shell
#!/bin/bash
......