Impala Public Jenkins has submitted this change and it was merged. ( 
http://gerrit.cloudera.org:8080/12681 )

Change subject: IMPALA-6326: part 2: remove fetch thread in stress test
......................................................................

IMPALA-6326: part 2: remove fetch thread in stress test

Ensure cursor is only accessed from a single thread. The means reworking
the code so that we check the time limit between fetch calls.

Use EXEC_TIME_LIMIT_S as an alternative to the previous multi-threaded
cancellation logic - it allows queries to be cancelled even when the
client is blocked or slow. This is implemented with the concept of
a CancelMechanism that determines *how* a query should be cancelled.
Query timeouts (where we want to cancel queries that run longer
than expected) are implemented using both cancel mechanisms, in
case the client is stuck in fetch or similar. Expected cancellations
are implemented with a random mechanism so that both code paths get
covered.

Testing:
Ran a cluster stress test.

Ran a couple of single-node stress tests with TPC-H and random queries.

Change-Id: If9afd74e1408823a9e5c0f2628ec9f8aafdcec69
Reviewed-on: http://gerrit.cloudera.org:8080/12681
Reviewed-by: Impala Public Jenkins <[email protected]>
Tested-by: Impala Public Jenkins <[email protected]>
---
M tests/stress/concurrent_select.py
M tests/stress/query_runner.py
2 files changed, 128 insertions(+), 90 deletions(-)

Approvals:
  Impala Public Jenkins: Looks good to me, approved; Verified

--
To view, visit http://gerrit.cloudera.org:8080/12681
To unsubscribe, visit http://gerrit.cloudera.org:8080/settings

Gerrit-Project: Impala-ASF
Gerrit-Branch: master
Gerrit-MessageType: merged
Gerrit-Change-Id: If9afd74e1408823a9e5c0f2628ec9f8aafdcec69
Gerrit-Change-Number: 12681
Gerrit-PatchSet: 9
Gerrit-Owner: Tim Armstrong <[email protected]>
Gerrit-Reviewer: David Knupp <[email protected]>
Gerrit-Reviewer: Impala Public Jenkins <[email protected]>
Gerrit-Reviewer: Thomas Marshall <[email protected]>
Gerrit-Reviewer: Tim Armstrong <[email protected]>

Reply via email to