-
Type:
Bug
-
Resolution: Duplicate
-
Priority:
Major - P3
-
None
-
Affects Version/s: 3.6.6
-
Component/s: Storage
-
Storage Execution
-
ALL
-
Storage NYC 2018-07-30, Storage NYC 2018-08-13, Storage NYC 2018-08-27, Storage NYC 2018-09-10, Storage NYC 2018-10-22, Storage NYC 2018-11-05
-
0
-
None
-
None
-
None
-
None
-
None
-
None
-
None
The update portion of the repro script repro-10MBx2.sh from SERVER-33191 shows ~10x regression between 3.6.5 (left) and 3.6.6 (right).
This seems to be related to eviction of large pages. Interestingly even though we are doing only about 9 updates per second, we are doing 170 evictions per second.
- depends on
-
WT-4483 Improve caching of small updates to large values
-
- Closed
-
- duplicates
-
WT-4483 Improve caching of small updates to large values
-
- Closed
-
- is related to
-
SERVER-31679 Increase in disk i/o for writes to replica set
-
- Closed
-
-
SERVER-33191 Cache-full hangs on 3.6
-
- Closed
-
-
SERVER-36373 create test to fill WT cache during steady state replication
-
- Closed
-
-
WT-4105 Optimize cache usage for update workload with history pinned
-
- Closed
-
- related to
-
SERVER-35958 Big CPU load increase (×4) on secondary by upgrading 3.4.15 → 3.6.5
-
- Closed
-