How To
Summary
During a system upgrade of Content Analyzer on a deployment that uses Operator, it can take an hour or more for the deployment to complete. This document describes how to reduce the deployment time.
Objective
Operator repeatedly loops through a list of tasks to install, upgrade, and modify pods. Depending on where it is in the loop when an upgrade request is received, Operator can be waiting for pods that are not upgraded yet. These pods will not come up running, so the task waits until it times out.
The upgrade is further complicated when there are multiple upgrades, such as when both Operator and Content Analyzer are upgraded. Operator eventually completes the upgrade without user intervention, but can require two or more cycles with corresponding waits each cycle until it gets everything properly upgraded.
If the administrator does not want to wait for the upgrade process to complete naturally, they can use this procedure to assure that Operator upgrades in the most optimal sequence, reducing the total upgrade time.
Steps
- Immediately before you apply the upgraded cr yaml files, scale down the Operator pod.
- oc scale deploy ibm-cp4a-operator --replicas=0
- Apply all cr yaml files as directed by your upgrade documentation.
- Scale back up the Operator pod so that it starts running tasks from the beginning of its list.
- oc scale deploy ibm-cp4a-operator --replicas=1
Document Location
Worldwide
[{"Business Unit":{"code":"BU053","label":"Cloud & Data Platform"},"Product":{"code":"SSUM7G","label":"IBM Business Automation Content Analyzer on Cloud"},"ARM Category":[],"Platform":[{"code":"PF025","label":"Platform Independent"}],"Version":"All Version(s)","Line of Business":{"code":"LOB45","label":"Automation"}}]
Was this topic helpful?
Document Information
Modified date:
04 May 2020
UID
ibm16205589