Search results: Spectrum LSF

Showing 35 of 2588

add more columns for job export

from job details view, we can export current in file, but the list of avail columns of the RTM settings -> job export view is limited compared to what we can display in the job detail view and also what's avail in the DB table. Can you please e...

9 months ago in Spectrum LSF / Operations/RTM 1 Future consideration

3 VOTE

Group jobs to same cluster when using bsub -cluster feature

Goal: Group together lsf jobs, "regression", in such a way that bsub -cluster feature directs all jobs in the group to the same cluster. Proposal: a) bsub submits all jobs in a job array to a single cluster bsub -cluster all -J "ArrayName[1-10]" /...

about 2 years ago in Spectrum LSF / Scheduling 1 Delivered

7 VOTE

Issue with data purging : Wastage of vertica license

As discussed in the PMR, due to bad design of data purging in PA 9, there are two issues 1. Wastage of vertica license i.e. at times data is retained in the DB even beyond data retention period defined in SYS_tables_to_parition. 2. PA admin has no...

over 10 years ago in Spectrum LSF 2 Delivered

1 VOTE

For each job, records the number of execution host CPU cores in the job's table

As we evaluate the utility of SMT2/4/8, we would like to be able to measure the performance difference of the applications at scale by recording the number of slots of the compute host for every job. This would achieved by adding a tinyint unsigne...

over 1 year ago in Spectrum LSF / Operations/RTM 1 Future consideration

1 VOTE

provide a mechanism to avoid manually inserting hash code into thold.php and settings.php for new thold alerts to view details.

When we add our own thold alerts, we must add manually the hash code recorded into a mysql table into thold.php and settings.php scripts to be able to display details. Can you please redesign the code to avoid these steps ?

almost 2 years ago in Spectrum LSF / Operations/RTM 1 Delivered

2 VOTE

Provide Access to Job Buckets for Analysis

Currently RTM contains quite a bit of information on current cluster state, but it provides limited information about the state of the scheduler, what job buckets exists, and the dispatch rate from these various buckets. Create an interface that p...

almost 4 years ago in Spectrum LSF / Operations/RTM 1 Not under consideration

2 VOTE

Add Memory Statistics to Daily Statistics

When I wrote the daily statistics interface years ago, I had very little comprehension of the concept of Memory Efficiency, and therefore did not add memory_reserved, and mem_used, and max_memory to the daily statistics tables. But in todays HPC, ...

almost 4 years ago in Spectrum LSF / Operations/RTM 2 Not under consideration

2 VOTE

Add Effective Memory Utilization to All Statistics Tables

RTM maintains a consistent set of ETL's across multiple dimensions. It is the same table structure for all of these ETL's. However, a key metric is missing from the ETL's and it's the effective memory utilization. In other words sum(max_memory) /s...

almost 4 years ago in Spectrum LSF / Operations/RTM 1 Not under consideration

2 VOTE

RTM to/from LSF communication of Psusp-requeue job notifcation

We have reproduced the issue and it's a timing issue with RTM because LSF does not provide exitinfo when job's pre-exec fails and requeued. Since RTM relies on exitinfo to put it in the correct table, the changes will have to come from LSF. This w...

about 5 years ago in Spectrum LSF / Operations/RTM 0 Not under consideration

3 VOTE

For more accurate pend time

Product: IBM Spectrum Computing AnalyticsVersion: 9.1.4Actor: SKhynix, Korea HQ Customer want to know how many server's slot was not used and application license was pended and what user could not run own jobs in how long time. To find some inform...

almost 8 years ago in Spectrum LSF 1 Not under consideration

Please enter your email address

FILTER BY CATEGORY