Skip to Main Content
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 Data & AI Roadmaps (http://ibm.biz/Data-and-AI-Roadmaps) - Use this site to view roadmaps for Data & AI products.

IBM Employees should enter Ideas at https://hybridcloudunit-internal.ideas.aha.io/


Status Future consideration
Workspace Spectrum LSF
Components Scheduling
Created by Guest
Created on Jan 19, 2022

Reporting both host-based pending reason and main pending reason

"LSF does not support reporting both host-based pending reason and main pending reason (about license)."

Let me provide a portion of the bsub statement from our case to exemplify the customer's dilemma.

bsub -R "rusage[mem=3000.00:duration=1m,md_rdcbase_lsi_g=1.00:duration=1m||md_rdcbase_lsi_l=1.00]" sleep 100

In the following bsub statement, the job submission should be completed when either conditions of the OR statments is satisfied.

Condition 1: mem=3000.00:duration=1m,md_rdcbase_lsi_g=1.00:duration=1m
Condition 2: md_rdcbase_lsi_l=1.00

However, the customer is confused because when the job is submitted and pends, bjobs -l <JOB_ID> shows the following message:
Job's requirements for resource reservation not satisfied (Resource: md_rdcbase_lsi_l);

$ bhosts -s md_rdcbase_lsi_g md_rdcbase_lsi_l
RESOURCE TOTAL RESERVED LOCATION
md_rdcbase_lsi_g 10.0 0.0 ALL
md_rdcbase_lsi_l 0.0 10.0 ALL

Based on the customer's original job submission, if either resources are available, the job should run. However, in reality, they are not only experiencing a lack of the latter resource (Condition 2), but they also have insufficient amounts of memory available for the job. So conclusively both Condition 1 and Condition 2 were not met.

"At that time, the pending reason is always for the last condition that is not satisfied. This is current LSF behaviour." Thus, only the following pending reason "Job requirements for reserving resources (md_rdcbase_lsi_l) not satisfied". is displayed to the user. For that reason, the customers were confused since they know that md_rdcbase_lsi_g is available for use. Since this was the case, customers perceived this to be an LSF scheduling issue.

So far we considered the following bsub option where Memory reservation is included for both resources:

bsub -R "rusage[mem=3000.00:duration=1m,md_rdcbase_lsi_g=1.00:duration=1m||mem=3000.00:duration=1m,md_rdcbase_lsi_l=1.00]" sleep 100

"but for pending reason about license,
License is a cluster-based resource. It is used by the whole cluster. If a license is not enough for a job, job cannot start even if many servers have enough memory. So job pending reason is for license if license is enough. Then LSF does not check memory resource.

Memory is a host-based resource. It is used by individual servers. A server does not have enough memory for the job, but other servers might have enough memory.

That's to say, when scheduling a job, LSF checks license first for the job. If license is enough, LSF continues to check memory. If license is not enough, mark pending reason to license related reason.

For the customer's job, the first condition includes memory and license requirements. When the first condition is not satisfied, LSF checks the second condition. If the second condition is not satisfied, the pending reason is for the second condition. This is LSF design behaviour."

In conclusion, the customer still cannot perceive the lack of memory to complete their jobs since they can only see the pending reason for insufficient number of licenses available. As this is the case, a drastic need for a logic to show pending reason for Condition 2 of the bsub rusage OR statement is evident to prevent further confusion for LSF users.

Needed By Month
1 MERGED

reporting both host-based pending reason and main pending reason

Merged
"LSF does not support reporting both host-based pending reason and main pending reason (about license)." Let me provide a portion of the bsub statement from our case to exemplify the customer's dilemma. bsub -R "rusage[mem=3000.00:duration=1...
about 1 year ago in Spectrum LSF / Scheduling 0 Future consideration