As with any production system, executing an upgrade for vCommander® works best when you employ careful planning and processes. Doing so ensures you have adequately notified users of expected outages, includes working with a staging or development system, understanding how to best interact with Embotics® Technical Support to overcome an obstacles, and have a known good configuration to which you can roll back in the event of unexpected issues.
Whenever upgrading any system, it’s best to let your users know in advance so they aren’t surprised by the service not being available when they need it. Typical vCommander upgrades take about an hour from start to finish if you are following the best practices outlined here, and keeping up to date.
Embotics recommends updating the Service Portal message of the day 36 hours to a week in advance of performing the upgrade to advise users.
- Under the Configuration menu, choose System Configuration.
- Switch to the Service Portal tab.
- Enter your message in the Service Portal Message of the Day field and click Save Settings.
Here’s a sample message:
To serve you better, a software upgrade has been scheduled for 17:00 – 18:00 Thursday, September 11, 2014. If you have any questions or concerns, or notice anything unusual once service has been restored, please let the system administrator know at firstname.lastname@example.org.
Additionally, send an email notifying users who may not login to the Service Portal during the period the message of the day is active.
Users rarely encounter vCommander upgrade failures, but in the few cases where failures occur, most often than not the problem lies with information in the database being in an unexpected format, or similar issues. As such, Embotics recommends maintaining a staging environment so you can perform a test run on upgrades prior to upgrading your production environment. Your vCommander license does permit the upkeep of a staging environment, but you should let your account manager and/or the support team know that you have one running. A permanent staging environment is also useful for trying out new workflows or policies, and becoming familiar with new vCommander features before implementing them as solutions in your production environment.
If you do not wish to maintain a persistent staging environment, you can also clone your app server and do a test run, if you are installed using the default postgres SQL database. Installations against SQL databases will require a bit more work, but can be similarly handled following the procedure below.
- Backup the SQL database.
- Create a new “dummy” database on your SQL Server.
- Install the version of vCommander you are currently running against the dummy database. This install will be upgraded as your test. Once you’ve completed the testing you can destroy the database.
- Stop the vCommander Windows service.
- Restore the backup of your SQL database, overwriting the dummy database.
- Restart the vCommander Windows service.
- Perform the upgrade.
In the event a vCommander upgrade fails, the installer will provide the location of a failure log that will provide details for Embotics support. Please do not run the installer a second time.
Along with emailing the support team, if you also have previously been provided ftp access to upload, it’s also a good idea to provide your backup of the database from before your upgrade attempt. That way, Embotics support and engineers have immediate access to investigate, develop and release a resolution to you.
Once you have successfully upgraded your vCommander, remember to complete the following activities:
- Turn off or update the Service Portal Message of the Day.
- Test the most common actions that your users employ when working with vCommander. If you experience any failures or unusual behavior, contact Embotics Support, providing as much detail as possible. This may include diagnostics packages, screenshots, or a backup of the upgraded database itself.
- Take database backups and/or snapshots of the vCommander application server.