-
Type:
Task
-
Resolution: Done
-
None
-
Affects Version/s: None
-
Component/s: None
Split out from WT-1250. This one, the plot and the chart from the AWS systems clearly show that the random code is not the problem for medium-lsm. The changes listed in the med-lsm plot between the good and bad are in build 1257; 21 diffs, mostly LSM compact, but a few things in there as well.