Spectrum LSF

Showing 391

possibility to distribute event files over many directories or subdirectories

If there are thousands of event files under 1 directory the speed to access the files will be slow and much faster if distributed over many directories

10 months ago in Spectrum LSF / Data Management 1 Not under consideration

1 VOTE

Multiple queues request for job - consider taking in account queue limitations and lsf.resources for better job dispatching.

Today, when we submit a job with multiple queues (bsub "high risk") The job will be dispatched to a relevant queue by examine parameters such as queue status(open/closed), queue hosts, time-limit. In a situation, where a queue "high" have a ujob_l...

10 months ago in Spectrum LSF / Scheduling 1 Not under consideration

5 VOTE

Set user limit for app in lsb.application like USERS field for queues in lsb.queues

It would be useful to set a list of users that can use APPS defined in lsb.applications like USERS filed il queues definition in lsb.queues.

about 4 years ago in Spectrum LSF / Administration 1 Not under consideration

2 VOTE

In the world of Multi-core processors, the context switches per second has become a KPI. It should be a standard performance metric like mem, r1m, etc.

When we try to oversubscribe hosts, it would be very beneficial to have ctx/second available as a base metric. It's available in all OS' these days. We could set a busy threshold for that say above 40k/second. For the longest time (maybe 31 years)...

over 1 year ago in Spectrum LSF / Administration 1 Not under consideration

1 VOTE

About /run/user/$uid

Could you please make "/run/user/$uid" the same for both ssh and interactive jobs? when user logon the host, /run/user/$uid is created by pam_systemd and used for storing files used by running processes for that user. When a user logs out, a...

12 months ago in Spectrum LSF / Administration 3 Not under consideration

2 VOTE

Enable option for hardcoding GPU memory request like its available for CPU memory request

At this moment IBM offers features of hardcoding CPU memory. We have implemented this feature in our cluster with esub scripting. Similarly we want to have same feature available for GPU memmory (gmem) to be hardcoded. I have raised ticket to our ...

almost 2 years ago in Spectrum LSF / Scheduling 2 Not under consideration

8 VOTE

Increase MAX_JOBID in PE environment - 4194303 too low

When using PE in LSF the maximum job ID is 4194303.We have already cycled through the range in 12 months and we'll likely do it twice this year.Having jobs with duplicate IDs causes a lot of problems with things like job accounting.

about 8 years ago in Spectrum LSF 1 Not under consideration

1 VOTE

Add support for GPU jobs in brun

Currently, the brun command can force a non-gpu job to run using various numbers of CPU's on one to many hosts, but this is not possible for GPU jobs. Add the ability to specify the GPU's or gpu_id's for a forced run gpu job to use, and follow the...

about 1 year ago in Spectrum LSF / Administration 1 Not under consideration

3 VOTE

Add IBM LSF RTM alert interval by weekly basis, currently it only has daily alert notice in RTM pop-up column.

IBM Spectrum LSF RTM only have hours, daily basis of alert pop-up column could choose. For management of current alert aggregation, We need IBM to add this feature enhancement into RTM fix patch. The features enhancement came from needs of...

about 3 years ago in Spectrum LSF / Operations/RTM 1 Not under consideration

4 VOTE

Set a limit for service class

It would be useful to set a limit in lsb.resources for Serive Class "SLA" in order give some minimum guarantee with SLA and set a maximium number of resorucese usable in this SLA.It would be excellent to indicate the limit with the number of nodes...

over 4 years ago in Spectrum LSF / Scheduling 1 Not under consideration

Please enter your email address

FILTER BY CATEGORY