This shows you the differences between two versions of the page.
Both sides previous revision Previous revision Next revision | Previous revision Next revision Both sides next revision | ||
cluster:103 [2011/12/21 14:39] hmeij [Submit] |
cluster:103 [2011/12/21 14:48] hmeij [Best Practices] |
||
---|---|---|---|
Line 132: | Line 132: | ||
==== Submit 2 ==== | ==== Submit 2 ==== | ||
- | On the back end compute nodes, unless specified, the job runs inside your home directory. | + | On the back end compute nodes, unless specified, the job runs inside your home directory. |
* new submission file with edits | * new submission file with edits | ||
- | * -n implies reserve job slots (cpu cores) for job (not necesssary, SAS jobs will always | + | * -n implies reserve job slots (cpu cores) for job (not necesssary, SAS jobs will always |
* -R reserves memory, for example, reserve 200 MB of memory on target compute node | * -R reserves memory, for example, reserve 200 MB of memory on target compute node | ||
- | * scheduler creates unique dirs in scratch by JOBPID for you, so we' | + | * scheduler creates unique dirs in scratch by JOBPID for you, so we' |
+ | * but now we must copy relevant to scratch dir and results back to home dir | ||
< | < | ||
Line 163: | Line 164: | ||
* you can monitor the progress of your jobs from greentail while it runs | * you can monitor the progress of your jobs from greentail while it runs | ||
+ | |||
< | < | ||
[hmeij@greentail sas]$ ll / | [hmeij@greentail sas]$ ll / | ||
Line 172: | Line 174: | ||
</ | </ | ||
+ | ==== Best Practices ==== | ||
+ | * You may submit as many SAS jobs as you like, just leave enough resources available for others to also get work done | ||
+ | * Because SAS submission are serial, non-parallel jobs your -n flag is always 1 | ||
+ | * Reserve resources if you know what you need, especially memory | ||
+ | * Use /sanscratch for large data jobs with heavy read/write operations | ||
+ | * Queue ehwfd is preferentially for Gaussian users and stay off the stata and matlab queues | ||
+ | * Write smart SAS code, for example, use data set indexes and PROC SQL (this can be your best friend) | ||
+ | * ... suggestions will be added to this page | ||
\\ | \\ | ||
**[[cluster: | **[[cluster: |