Details
-
Bug
-
Status: Closed
-
Blocker
-
Resolution: Fixed
-
0.20.4, 0.20.5, 0.90.0
-
None
-
None
-
Reviewed
Description
While doing a concurrent read/write test, the reader eventually gets a situation where the first column in the result set has the wrong 'value' than the rest of the result set (of 50 columns or so). The test (included) does puts of 50 columns with all the same (Random) value. The reader validates that all values are equal, and fails.
Attachments
Attachments
Issue Links
- is related to
-
HBASE-2670 MemStore should retain multiple KVs with the same timestamp when memstoreTS differs
- Closed
-
HBASE-2616 TestHRegion.testWritesWhileGetting flaky on trunk
- Closed