VMware Infrastructure Health MP alerts and symptoms are missing from Aria Operations 8.18.x

Products

• VMware Aria Suite

Issue/Introduction

The VMware Infrastructure Health MP alerts and symptoms definitions are missing from Aria Operations 8.18.x

Environment

Aria Operations 8.18.x

Resolution

To resolve the issue, please follow the steps below.
Please note: The steps will clean up the orphaned SymptomRelationship rows.

• Take the cluster offline
• Take snapshots for all nodes or follow the steps on How to take a snapshot of Aria Operations
• Start the repl db on the Primary node
  


  /etc/init.d/vpostgres-repl start

  
  
  
• Dump the kv_symptomproblemdefinition table to a CSV
  


  su - postgres -c "/opt/vmware/vpostgres/current/bin/psql -d vcopsdb -p 5433 -c \"COPY kv_symptomproblemdefinition TO '/tmp/kv_symptomproblemdefinition_$(date +%Y%m%d).csv' WITH (FORMAT CSV, HEADER TRUE);\""
  zip /tmp/kv_symptomproblemdefinition_$(date +%Y%m%d).zip /tmp/kv_symptomproblemdefinition_$(date +%Y%m%d).csv

• Verify orphaned SymptomRelationship-SymptomDefinition-% rows still exist
  


  su - postgres -c "/opt/vmware/vpostgres/current/bin/psql -d vcopsdb -p 5433 -c \"
  SELECT
      sr.key
  FROM kv_symptomproblemdefinition AS sr
  WHERE sr.key LIKE 'SymptomRelationship-SymptomDefinition-%'
      AND NOT EXISTS (
          SELECT *
          FROM kv_symptomproblemdefinition AS sd
          WHERE sd.key = RIGHT(sr.key, -LENGTH('SymptomRelationship-'))
      );
  \""

  
  
  
• Delete the orphaned SymptomRelationship-SymptomDefinition-% rows
  


  su - postgres -c "/opt/vmware/vpostgres/current/bin/psql -d vcopsdb -p 5433 -c \"
  DELETE FROM kv_symptomproblemdefinition AS sr
  WHERE sr.key LIKE 'SymptomRelationship-SymptomDefinition-%'
      AND NOT EXISTS (
          SELECT *
          FROM kv_symptomproblemdefinition AS sd
          WHERE sd.key = RIGHT(sr.key, -LENGTH('SymptomRelationship-'))
      );
  \""

  
  
  
• Verify no orphaned rows remain
  


  su - postgres -c "/opt/vmware/vpostgres/current/bin/psql -d vcopsdb -p 5433 -c \"
  SELECT
      sr.key
  FROM kv_symptomproblemdefinition AS sr
  WHERE sr.key LIKE 'SymptomRelationship-SymptomDefinition-%'
      AND NOT EXISTS (
          SELECT *
          FROM kv_symptomproblemdefinition AS sd
          WHERE sd.key = RIGHT(sr.key, -LENGTH('SymptomRelationship-'))
      );
  \""

  
  
  
• Bring cluster online
• Perform one of the following:
  • Re-install ONLY the VIH MP.  On the PRIMARY node, SSH and run the following API command to re-install VMwareInfrastructureHealth-818XXXXX.pak, then watch the pak install progress on the "Admin UI > Software Update" page.
    


    curl -ksS -w "%{http_code}" -X POST -u "admin" -H "Accept: application/json" "https://localhost/casa/upgrade/cluster/pak/VMwareInfrastructureHealth-818XXXX/operation/reinstall"

    Please note: It will be required to find the minor version and the build number for the VMware Infrastructure version that is suitable for your environment. The information can be found in the path: /storage/db/pakRepoLocal/ Filename: VMwareInfrastructureHealth-818[minor version][build number] - Make sure to pick the most recent/highest build number of the VMware Infrastructure Health MP.

Check for the existence of expected Infrastructure Health MP alerts and symptoms in the Product UI