NPE when listing repairs about cassandra-reaper HOT 4 CLOSED

Which lambda are we wanting to continue here? I'm not sure what you're referring to. Am I right in thinking that we want to check that the ResultSet returned by row.getUninterruptibly() is not empty?

If the ResultSet is empty, do we want to throw a ReaperException for handling up the stack, or just log and return an empty List<RepairRun>?

(NB: I'll tackle this since it is surely an error in code I've written...)

from cassandra-reaper.

I'll de-prioritize this for now, the root cause was a consistency issue in the db.
This is a nice to have but not high priority.

from cassandra-reaper.

I would have been able to work on this today (as I think my other tickets are done, and I don't have anything else queued up) but I need answers to the above questions @adejanovski...

from cassandra-reaper.

This hit us as well, and I ended up making a PR before noticing the Issue: #1478

DB inconsistency is the root cause, but in our case it was cased by Reaper filling the disk with logs (which it tends to do if anything goes wrong) on the VM. Enough that I think we've hit it a few times before, but just solved it previously by wiping the DB. This round I had time to go digging.

from cassandra-reaper.