OSCER-USERS-L Archives

OSCER users

OSCER-USERS-L@LISTS.OU.EDU

Options: Use Classic View

Use Proportional Font
Show HTML Part by Default
Show All Mail Headers

Topic: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Henry Neeman <[log in to unmask]>
Mon, 9 May 2022 10:35:54 -0500
text/plain (59 lines)
OSCER users,

REMINDER:

Because of Schooner's scheduled maintenance outage this coming
Wed (May 11), if you submit batch jobs to Schooner between now
and then, it might happen that those jobs get delayed until
after the scheduled maintenance outage ends.

You can get such jobs to run before the scheduled maintenance
outage by reducing the amount of wall clock time you request,
specifically the following #SBATCH directive:

#SBATCH --time=DD-HH:MM:SS

OR

#SBATCH -t DD-HH:MM:SS

DD: 2-digit number of days
HH: 2-digit number of hours
MM: 2-digit number of minutes
SS: 2-digit number of seconds

(Or you might use a different time format, like HH:MM:SS
without the number of days and the hyphen.)

For example, if you submit the job at noon today, be sure to
request at most 44 hours (you should probably request 40 hours
or less, to have a margin of error).

We apologize for the inconvenience -- our goal is to make
OSCER systems and services even better!

The OSCER Team ([log in to unmask])

----------

On Thu, 5 May 2022, Henry Neeman wrote:

>OSCER users,
>
>This coming Wednesday (May 11) 8:00am-midnight, we'll hold
>an outage to upgrade to the most recent stable version of
>the Slurm batch system (sbatch, squeue, etc), to resolve
>a security vulnerability in older versions of Slurm.
>
>During the outage, batch jobs won't be able to run.
>
>This means that, on Monday and Tuesday, if you want a job to
>run, you'll need to make sure that it'll be able to end
>by Wednesday 8:00am (so you might need to reduce its
>time limit).
>
>We apologize for the inconvenience -- our goal is to make
>OSCER systems and services even better!
>
>The OSCER Team ([log in to unmask])

ATOM RSS1 RSS2