Bharath Vissapragada has posted comments on this change.

Change subject: IMPALA-3680: Cleanup the scan range state after failed hdfs 
cache reads
......................................................................


Patch Set 3:

(2 comments)

Yes the test caught the wrong offset bug. Due to the wrong offset, the no. of 
rows scanned from cached_orders were more than the original.

http://gerrit.cloudera.org:8080/#/c/3313/3/be/src/runtime/disk-io-mgr.cc
File be/src/runtime/disk-io-mgr.cc:

Line 535:         range->Close();
> i think it would be better to put this in the ReadFromCache() function so t
Actually I tried that first but it creates a deadlock on hdfs_lock_. Both 
Close() and ReadFromCache() try to acquire it. So I moved it to the callers.


http://gerrit.cloudera.org:8080/#/c/3313/3/tests/query_test/test_hdfs_caching.py
File tests/query_test/test_hdfs_caching.py:

Line 118:       self.execute_query("ALTER TABLE tpch.cached_orders set cached 
in 'testPool'")
> we may need to skip this on S3/isilon/dssd.
Done


-- 
To view, visit http://gerrit.cloudera.org:8080/3313
To unsubscribe, visit http://gerrit.cloudera.org:8080/settings

Gerrit-MessageType: comment
Gerrit-Change-Id: I0a9ea19dd8571b01d2cd5b87da1c259219f6297a
Gerrit-PatchSet: 3
Gerrit-Project: Impala
Gerrit-Branch: cdh5-trunk
Gerrit-Owner: Bharath Vissapragada <[email protected]>
Gerrit-Reviewer: Bharath Vissapragada <[email protected]>
Gerrit-Reviewer: Dan Hecht <[email protected]>
Gerrit-HasComments: Yes

Reply via email to