perfSONAR Announcements

[Mark Feit <>]

Following up on my earlier message:

 

Systems that failed as a result of the upgrade to 5.2.1 can be brought into a functioning state by logging in, becoming root and executing the following commands:

  

• pscheduler internal db-change-password
• pscheduler internal warmboot
• pscheduler internal service restart
• pscheduler troubleshoot

 

The troubleshoot  should complete successfully with the last couple of lines looking like this (the number of seconds may be different):

 

  Idle test.... 4 seconds... Finished... Checking archiving... OK.

 

pScheduler appears to be functioning normally.

 

If the troubleshoot fails because of a service that isn’t running, repeat the last two steps and things should start working.  Please drop me a line if it fails for any other reason.

 

This will leave you with a system that includes all of 5.2.1’s changes except one minor change that corrects what the pScheduler troubleshooter displays as the last-completed run.   The rest of what’s in 5.2.1 can be found in the release notes:   https://www.perfsonar.net/releasenotes-2025-07-01-5-2-1.html.

 

The repositories are still as described previously.  I’ll follow up again once they’ve been updated.  This should not have had any effect on the Docker container.

 

Special thanks to Ricard Cruz at PIC for testing out this fix before we released it.

 

--Mark