-
Type: Improvement
-
Resolution: Done
-
Priority: Major - P3
-
Affects Version/s: None
-
Component/s: None
-
None
A while ago, keith.bostic asked me:
In __wt_connection_workers, why are we doing recovery so late? Specifically, we're starting the async threads, and the statistics logging thread?
Why not let recovery do its stuff as soon as possible?
There does not appear to be any particular reason for recovery to be run that late. I'll look at moving it earlier and test that out. I agree running before some of the other service threads makes sense.
- is depended on by
-
SERVER-18199 WiredTiger changes for MongoDB 3.1.2
- Closed