16 Sep
2011
16 Sep
'11
2:22 a.m.
Johan Meiring wrote:
It was just the first time that I didn't see FR "recover" after the DB eventually finished with slow queries.
I made a minor change for 2.1.12 which will clear out the pending queue a lot more quickly. old: for each new request clear out one "dead" request from queue If the dead requests originally arrived at a higher rate than the new requests, it can take a long time to clean them out. new: for each new request clear out all "dead" requests from queue Much better. Alan DeKok.