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 updateson 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,
Post an idea
Upvote ideas that matter most to you
Get feedback from the IBM team to refine your idea
Specific links you will want to bookmark for future use
hostcache auto-delete function in the mbatchd restart
Dynamic hosts remain in the cluster unless you intentionally remove them.And the manual related to it is as follows.https://www.ibm.com/docs/en/spectrum-lsf/10.1.0?topic=cluster-remove-dynamic-hostsSuppose we delete the hostcache file, not modify ...
Have LSF Handle ZOMBI job searching using either BTREE or HASH ALGO's
Large number of ZOMBI jobs can impact very large clusters operations resulting in high backlogs and additional lost job status. Improving the ZOMBI search algorithm from Linear Search to BTREE or HASH would reduce this impact.
Enhance the Support Tool to be able to remove ZOMBI jobs and create a log of exec hosts and pgids that need to be cleaned
We had an issue where LSF could not keep up with dispatch due to high numbers of ZOMBI jobs. We would like the ability to shut down LSF and clean up the ZOMBI jobs via a Database repair to accelerate the time-to-restore of the LSF clusters.
Make Queue Open/Close/Act/Inact 'all' Requests a single event
We recently had a serious issue that required us to close/inact all of our queues, and then perform a database repair using the IBM support tool. However, when we started LSF, all but one of the queues was Open/Act again. This was due to the suppo...
Make the TCP Backlog for the MBATCH and MBATCHD Query Port Configurable for Large Clusters
We have been monitoring the TCP backlogs for LIM, MBD, and the Query port for some time, and there are periods where the MBD query port becomes overloaded resulting in a higher than expected count of UNKNOWN and sometimes eventually ZOMBI jobs, an...
Add lsb.params setting to limit the number of times a job can be redispatched upon either a pre-exec failure of job initilization failure
We currently have jobs that will bounce from host to host either based upon a pre-exec failure, or a job initialization failure. This can impact > 500 hosts and cause an increase in MBD memory use. It would be nice to have a setting such as: MA...
Improve Job Priority Output to Exceed 1e6 as APS can easily increase Per Job Priority above that value
There is currently a system limitation that Job Priories above 1e6 do not display correctly when using bjobs -o "blah" formatted output. This is currently causing us a monitoring issue when attempting to order jobs with a very high dynamic priorit...
Support Multi-threaded LIM with a Shared Memory Database for Scalability
It would be nice if LSF's LIM used a shared memory database and could accept multiple threads instead of having to constantly fork itself to return client data. The new setting to increase/decrease the LIM connection backlog was informative, but i...
Do not place IBM confidential, company confidential, or personal information into any field.