Description
I just pulled a recent repo sync and running a dev build (osx), with moss commit 0bce41eed28, I see cbft consuming 500% cpu after a restart. Not sure how I got into this state, but I had 7 FTS default indexes on the beer-sample bucket. And, from repeated stack dumps, I see that moss is just seemingly infinite looping in calcPartialCompactionStart() / determineExponent(). Please see attached stack dump.
Perhaps the default CompactionLevelMultiplier of 1 billion overflows? Perhaps the determineExponent should check for overflow? Perhaps determineExponent should have a max exponent param where it gives up?
Attachments
Issue Links
- is duplicated by
-
MB-24978 [FTS] Rebalance fails during indexing with bad_nodes,fts,get_node_infos error
- Closed