IBM Data and AI Ideas Portal for Customers


This portal is to open public enhancement requests against products and services offered by the IBM Data & AI organization. To view all of your ideas submitted to IBM, create and manage groups of Ideas, or create an idea explicitly set to be either visible by all (public) or visible only to you and IBM (private), use the IBM Unified Ideas Portal (https://ideas.ibm.com).


Shape the future of IBM!

We invite you to shape the future of IBM, including product roadmaps, by submitting ideas that matter to you the most. Here's how it works:


Search existing ideas

Start by searching and reviewing ideas and requests to enhance a product or service. Take a look at ideas others have posted, and add a comment, vote, or subscribe to updates on them if they matter to you. If you can't find what you are looking for,


Post your ideas

Post ideas and requests to enhance a product or service. Take a look at ideas others have posted and upvote them if they matter to you,

  1. Post an idea

  2. Upvote ideas that matter most to you

  3. Get feedback from the IBM team to refine your idea


Specific links you will want to bookmark for future use

Welcome to the IBM Ideas Portal (https://www.ibm.com/ideas) - Use this site to find out additional information and details about the IBM Ideas process and statuses.

IBM Unified Ideas Portal (https://ideas.ibm.com) - Use this site to view all of your ideas, create new ideas for any IBM product, or search for ideas across all of IBM.

ideasibm@us.ibm.com - Use this email to suggest enhancements to the Ideas process or request help from IBM for submitting your Ideas.

IBM Employees should enter Ideas at https://ideas.ibm.com


ADD A NEW IDEA

Scheduling

Showing 46

for loadSched memory if the submitted job has memory reservation the job should not be run if the nodes mem used + job memory reservation > loadSched memory limit

because users want to keep a certain amount of memory free. And currently the limit can be bypassed.
about 18 hours ago in Spectrum LSF / Scheduling 0 Submitted

Effective r15s/r1m/r15m values for job reservations

When using run queue length indices in rusage strings, there should be an option to have the value divided by ncpus. Since in the "total load used for scheduling" the r15s/r1m/r15m are taken from lsload -E and divided by the specific ncpus for tha...
21 days ago in Spectrum LSF / Scheduling 2 Not under consideration

Affinity scheduling must take L3 cache hierarchy into account

Newer CPUs have multiple L3 caches, with dedicated cores per L3 cache. Running multi-threaded jobs on all cores within a shared L3 cache can reduce runtime in our cases up 70%, compared to running over multiple L3 caches. Think of the AMD EPYC 935...
about 2 months ago in Spectrum LSF / Scheduling 2 Planned for future release

Map memory reservation to cgroup memory.low

Mapping the memory reservation of a job to the cgroup memory.low could protect jobs from running short on memory, while in parallel allowing more memory to be used if it's available. The soft limit instead of a hard limit should prevent users to b...
about 2 months ago in Spectrum LSF / Scheduling 2 Future consideration

Keep job in memory for a specified time period

Allow job to be kept in memory for an extended amount of time beyond global setting
about 2 months ago in Spectrum LSF / Scheduling 0 Future consideration

command-level job starters should work with bsub

According to https://www.ibm.com/docs/en/spectrum-lsf/10.1.0?topic=starters-about-job command-level job starters (using LSF_JOB_STARTER env variable) do not work Are user-defined. They run interactive jobs submitted using lsrun, lsgrun, or ch. Com...
2 months ago in Spectrum LSF / Scheduling 5 Is a defect

Improve bsub rejection messages for troubleshooting

Improve the error messages returned from bsub failiure. For example, provide the rusage string and potentially specified what is wrong with rusage. Here is current failure message: Error in rusage section. Job not submitted.
2 months ago in Spectrum LSF / Scheduling 0 Planned for future release

Add job group filter to bhist

The ability to find job groups via bhist command after the jobs have been removed from memory.
2 months ago in Spectrum LSF / Scheduling 0 Future consideration

LSF job query filter on pending jobs without providing pending reasons

For large LSF clusters, the bjobs command takes significant time to gather the pending reasons (-p, -pei, -pe, -pi) making the query much longer than requesting all jobs. Suggest a new bjobs option to filter on pending jobs without supplying the p...
3 months ago in Spectrum LSF / Scheduling 0 Planned for future release

Add possibility to assign cores from the same CCX in AMD CPUs to make use of the same L3 cache.

With a specific 2-core workload we noticed significant performance improvements of about 50% in shorter runtime when both cores reside on the same L3 cache group (CCX) of an AMD EPYC CPU.Therefore it would be very beneficial to be able to have LSF...
over 1 year ago in Spectrum LSF / Scheduling 0 Planned for future release

By clicking the "Post Comment" or "Submit Idea" button, you are agreeing to the IBM Ideas Portal Terms of Use.
Do not place IBM confidential, company confidential, or personal information into any field.