Best Practices Guide prior to Release Automation upgrade
search cancel

Best Practices Guide prior to Release Automation upgrade

book

Article ID: 40617

calendar_today

Updated On:

Products

CA Release Automation - Release Operations Center (Nolio) CA Release Automation - DataManagement Server (Nolio)

Issue/Introduction

Consideration and best practice recommendation for Release Automation upgrade

Release Automation is configured with various integration configurations and there are some guidelines before upgrading. The guidelines will assist for successful upgrade.

Glossary

  • DMS a.k.a NAC: Data Management Server /Nolio Automation Center

  • ES a.k.a NES: Execution Server / Nolio Execution Server

  • Agents

Environment

Release Automation: 6.4, 6.5, 6.6 and higher

Resolution

Upgrade Checklist

  • Keep the service account username and password detail running DMS and ES services handy before running the upgrade.
  • Before the upgrade, make a backup of files in <InstallRootDir>/conf  and cacerts file in <InstallRootDir>/jre/lib/security.
  • Make sure the user whose account is used for upgrade has adequate permissions on DMS and ES and install directories.
  • Jobs:
    • Make sure there are no active jobs or hanged jobs on DMS. Stop all of them before upgrade
    • Make sure there are no scheduled jobs and Test run running. See Appendix A on how to do it.
  •  In High Availability setup, make sure that when the upgrade is running, neither Primary/Secondary DMS are processing any request. Both the primary and secondary server require the patch in specific order. During patch installation, there are no database updates other than the product version. Only the Management Server receives updates to required files.
    1. Stop the secondary Management Server
    2. Apply the cumulative patch on the Primary Management Server. The patch installation stops the service, updates the required files and restart the services.
    3. Verify that the primary Management Server is available and working properly. 
    4. Apply the cumulative patch on the secondary Management Server. The patch installation stops the service, updates the required files and restart the services.
Note: There can be various ways in which HA on DMS is set up - it may be via Load balancers, VIP, etc. An important thing to remember is that during the upgrade, the primary DMS will go down for some time during the restart.
  • Make sure that the secondary does not become primary and thus tried to communicate with the database. This may result in DB tables getting locked and upgrade failures.
    • So, make sure that secondary Management Server is stopped during the upgrade process.
  • Create back up of necessary certificates like of LDAP certificates.

Additional Information

Appendix A

Seeking Active Run: There are four possible executing objects in the systems

  1. Releases

  2. Processes

  3. Scheduled Jobs

  4. Test Run

 Checking Active Releases

  1. Navigate to JMX URL http://<DMS-Hostname/IP>:20203/mbean?objectname=releasecenter%3Atype%3DInfo and enter JMX login details.
  2. Invoke method getAllActiveReleases, it will list down all active releases. The desired result expected before any upgrade/maintenance on Server is to have “Result value: No active releases found”.
  3. If there are no active releases, you can skip this process. If you have some releases, either stop them via Release Operation Center UI or you can invoke method removeRelease passing the release-id from URL in step-1.

Checking Active Processes

  1. Navigate to JMX URL http://<DMS-Hostname/IP>:20203/mbean?objectname=noliocenter%3Atype%3Dinfo and enter JMX login details.
  2. Invoke method listActiveRuns, it will list down all active process runs. The desired result expected before any upgrade/maintenance on Server is to have “Result value:”.
  3. If there are no active runs, you can skip this process. If you have any active runs, either remove them all by invoking method removeAllActiveRuns or remove them in one-by-one by invoking removeProcessRuns passing process-id’s. Both methods can be accessed navigating to URL in step 1.

Checking Scheduled Jobs

  1. Navigate to JMX URL http://<DMS-Hostname/IP>:20203/mbean?objectname=noliocenter%3Atype%3Dinfo and enter JMX login details.
  2. Invoke method listAllScheduledJobs, it will list down all scheduled jobs. The desired result expected before any upgrade/maintenance on Server is to have "Result value: No scheduled jobs", but it may not be possible.
    1. If there are any scheduled jobs, you will get the list of jobs with three column information  i.e. ID, NAME,ENABLED.
    2. You can either disable these schedules from RA if scheduled jobs are conflicting with the maintenance window(Management Server upgrade time frame)
  3. If you want to remove a scheduled job you think will fall under method removeScheduledJobs passing scheduled-id captured in step 2. This method can be accessed navigating to URL in step 1.

Removing Test Runs

  1. Navigate to JMX URL http://<DMS-Hostname/IP>:20203/mbean?objectname=noliocenter%3Atype%3Dinfo and enter JMX login details.
  2. Invoke method removeTestRuns it will remove any test runs been there in system. Test runs are one which we invoke by executing test on process/actions.
Powered by Wolken Software