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:
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
Help IBM prioritize your ideas and requests
The IBM team may need your help to refine the ideas so they may ask for more information or feedback. The product management team will then decide if they can begin working on your idea. If they can start during the next development cycle, they will put the idea on the priority list. Each team at IBM works on a different schedule, where some ideas can be implemented right away, others may be placed on a different schedule.
Receive notification on the decision
Some ideas can be implemented at IBM, while others may not fit within the development plans for the product. In either case, the team will let you know as soon as possible. In some cases, we may be able to find alternatives for ideas which cannot be implemented in a reasonable time.
Block hosts per resource status or add hosts blocking in Notebook configuration
Conductor failing to start notebooks remain in tentative as it is unable to detect the host mem/cpu usage. Need the feature in Notebook configuration to block hosts per resource status or add hosts blocking in Notebook configuration.
Conda Package for Pytorch - ppc64le with Cuda 10.2
We are currently using Spectrum Conductor 2.5 on RHEL 7.6 with Cuda 10.2 on Power 9 servers. The model validators need the latest Pytorch Package to validate models with the latest features. The current IBM repo channels are having packages that a...
We are provide permission at group level ,when we add new users to group, user level permission it is not synched. bash-4.2$ egosh user roles4user -u u003464 No role for user <u003464>. bash-4.2$ getent group cmrddsairaw cmrddsairaw::3059:a0...
We have a requirement where we need to set a cap at the sig level so that no application can go above on the maximum number of slots. Use case===========At the SIG level I have set the SPARK_EGO_SLOTS_MAX: 10 which is the max allowable for any use...
Run SparkCleanup service per individual Instance group instead of centrally
We would like SparkCleanup service to run as execution user of each individual Instance Group, rather than centrally as cluster admin. This will avoid the need to do preliminary work to make sure execution users of SIG are part of cluster admin un...
Currently Spark Master runs as the execution user specified in the SIG which happens to be different from cluster admin and depends on SIG tenant/consumer group. That spark master resources such as logs are owned by tenant user makes it difficult ...
Optimize SIG package deployment for upgrade or rollback
SIG package deployment takes upto 10 minutes or so in a 50+ SIGs and 1000+ nodes cluster. Refactoring SIG deployment such that package distribution and enablement are independently managed. This will allow scheduling new SIG package distribution a...
Do not place IBM confidential, company confidential, or personal information into any field.