Details
-
Bug
-
Resolution: Fixed
-
Major
-
1.6.5
-
None
-
Security Level: Public
-
None
-
EC2 CentOS
Description
I created a 3 node cluster with 6M items and 2k item size. My working set was 1M and I had EC2 m1.large instances. When I did a failover and then rebalance the performance on my cluster dropped significantly. I went from doing 3.5k ops/sec to doing 600 ops/sec with many drops below 100 ops/sec. Attached is a screenshot of performance after the rebalance. Also performance slowly increases over time until the rebalance finishes.