TrxRegionEndpoint coprocessor performScan error

[Joanie Cooper]

During the execution of the SQL regression test, catman1/TEST137, an occasional HBase error is returned.

55> Unexpected EXCEPTION : *** ERROR[8448] Unable to access Hbase interface. Call to ExpHbaseInterface::nextRow returned error HBASE_ACCESS_ERROR(-705). Cause:
java.util.concurrent.ExecutionException: java.io.IOException: PerformScan error on coprocessor call, scannerID: 0
java.util.concurrent.FutureTask.report(FutureTask.java:122)
java.util.concurrent.FutureTask.get(FutureTask.java:188)
org.trafodion.sql.HBaseAccess.HTableClient.fetchRows(HTableClient.java:419)
. [2015-01-29 06:07:47]

Analysis has shown that a TrxRegionEndpoint coprocessor "openScanner" API call error can be reported as a TrxRegionEndpoint coprocessor "performScan" error.

The TrxRegionEndpoint coprocessor "performScan" API call error can also be returned when SQL performs a multi-threaded prefetch operation.

The TrxRegionEndpoint coprocessor "performScan" API calls can be executed out of sequence causing scanner sequence numbers to
not match. The coprocessor returns the error, as the client/server scanner sequence numbers must match.

[Atanu Mishra (atanu-mishra)]

Per Selva --

If a scanner is closed before the next call completes, it results in OutOfOrderScannerNextException. Trafodion engine can close the scanner when the next is active in a different thread with pre-fetch concept. The fix is to cancel the pending next before the scanner is closed.

Thanks to Oliver and Joanie for tracking this down. Ran catman1/TEST137 many times successfully

[Selvaganesan Govindarajan (selva-ganesan)]

Fixed committed See https://review.trafodion.org/1135