Details
-
Bug
-
Status: Closed
-
Major
-
Resolution: Fixed
-
None
-
None
-
None
Description
A region move/split/rebalance should be delayed when a TransactionalScanner is present on the region coprocessor 'scanners' list. The delay is not happening and is throwing the following error:
-
-
- ERROR[8448] Unable to access Hbase interface. Call to ExpHbaseInterface::nextRow returned error HBASE_ACCESS_ERROR(-706). Cause:
java.util.concurrent.ExecutionException: java.io.IOException: PerformScan error on coprocessor call, scannerID: 0 java.io.IOException: performScan encountered Exception txID: 562962838798562 Exception: org.apache.hadoop.hbase.UnknownScannerException: ScannerId: 0, already closed?
- ERROR[8448] Unable to access Hbase interface. Call to ExpHbaseInterface::nextRow returned error HBASE_ACCESS_ERROR(-706). Cause:
-
While debugging I noticed that in the SplitBalanceHelper the code is waiting for clear scanners list && pending txn list. It should be '||' instead.