Disconnected queues - cfg_job_status_queue and policy_managment since upgrade to 20.3.3

Products

• DX Unified Infrastructure Management (Nimsoft / UIM)
• Unified Infrastructure Management for Mainframe
• CA Unified Infrastructure Management SaaS (Nimsoft / UIM)

Issue/Introduction

Since the migration of UIM to 20.3.3 the cfg_job_status_queue and policy_managent queues are switched off, is this correct, are they needed for any probe, can they be disabled?

Environment

Release : 20.3

Component : UIM - ALARM POLICY

Resolution

cfg_job_status_queue is used for the ace probe which is no longer used (deprecated) in 20.3, so you can disable the ace probe if its still present and/or deactivate the queue and delete it.

policy_management is used for MCS so it should be ENABLED and green but if you're not using MCS the queues will remain at 0.

We create the default policy when an MCS enhanced profile is created. 

During profile creation we send the pds to the policy_management_queue with subject of "MCS_Profile_Event".

policy_management_ws (present in ump/oc)  subscribes to this queue and when it receives the pds with the above subject, the pds is processed to create the "default policy." 

Please download and deploy the latest hotfix -> policy_management-ws-0.2.7-HF2 from https://support.broadcom.com/download-center/solution-detail.html?aparNo=LU02691&os=MULTI-PLATFORM

Better yet, upgrade to the September release (monthly patch) (or October patch) if you haven't already.

Steps to apply the patch:

1. Import the zip into the local archive.
2. Deploy the imported package to the robot's running policy_management_ws

Additional steps:
a. If the policy_processing flag is set in wasp.cfg > webapps/policy_management_ws/custom/uncrypted/policy_processing to true, policy processing will happen, if set to false, processing will not happen. This is the manual way of configuration.

b. In a 'load balanced' environment with multiple policy_management_ws nodes, the policy_processing should happen in an HA mode. For that, webapps/adminconsoleapp/custom/uncrypted/ha_mode should be set to true and all the policy_management_ws nodes should not have the policy processing flag. Then, nodes will be selected dynamically and work in HA mode.

c. In a https environment, please add the configuration parameter->

controller_url = https://FQDN/Certname:<port>/adminconsoleapp

Via the wasp.cfg -> webapps/policy_management_ws/custom/unencrypted, to avoid certificate issues.

Additional Information

Related KBs:

MCS Alarm Policies management and policy_management queue (broadcom.com)

The policy_management queue on the primary hub is yellow / not connected (broadcom.com)

UIM - policy_management queue in secondary hub probe is in yellow

UIM - policy_mangement queue is yellow and no processing after deployment of java_jre 2.09 (broadcom.com)