Open bgrounds opened 5 years ago
This has been discussed inside the project before and I'd swear we had a ticket open on it but I can't find it. I guess this will be it.
This won't be difficult to do since the information about the run that would contain the diagnostics is already there.
Mental notes for later:
[ ] Need to start an archiving for runs where the status is considered final. This should be governed by a new column in run_state
and needs to exclude states like overdue
, which eventually becomes missed
(which is final). (Archivings are inserted by archiving_run_after()
in server's database/archive.sql
.)
[ ] Should probably add a Boolean archive_failures
to the archive spec to maintain existing behavior as the default.
Currently (as far as I can tell) I have to log into each of my perfSONAR hosts and manually check various log files and the pscheduler schedule for troubleshooting and investigation.
Test results from all of my hosts are available in a central dashboard, but I can't see, for example, which tests were Non-starters in this central dashboard, or any errors that may have occurred in any of perfSONAR's component services.
I'd like to be able to process a stream of errors (from various perfSONAR components) and scheduling metadata (e.g. Non-Starters) with the archiver.
note: I'm using a custom archiver to send the events to a third-party dashboarding solution