Details
-
Bug
-
Status: Closed
-
Major
-
Resolution: Fixed
-
0.3
-
None
-
None
Description
HBaseStore flush needs better synchronization:
Gora client threads have own HTable each. One thread can trigger a flush for every HTable. On top of that HTable has internal flushing. (When the write buffer is exceeded).
We have been running a patch that fixed this. Will attach and commit.