[
https://issues.apache.org/jira/browse/IMPALA-9090?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16978773#comment-16978773
]
ASF subversion and git services commented on IMPALA-9090:
---------------------------------------------------------
Commit 6b6e15b08ef799bbbb916c553b013dcd88ffb055 in impala's branch
refs/heads/master from xiaomeng
[ https://gitbox.apache.org/repos/asf?p=impala.git;h=6b6e15b ]
IMPALA-9090: Add name of table being scanned in scan node profile
Before this change, the only way to figure out the table being scanned
by a scan node in the profile is to pull the string out of the explain
plan or execsummary. This is awkward, both for manual and automated
analysis of the profiles. We should include the table name as a string
in the SCAN_NODE implementation.
After this change, a new line "Table Name: database.table" will be added
in first line of scan node profile.
Also fix a bug that frontend pass hbase and kudu table name incorrectly
to thrift. Before this change, native name of hbase and kudu table are
passed in and there is no way to get hms table name from TableDescriptor
in backend.
After this change, for HBaseTableDescriptor and KuduTableDescriptor,
function name() would return hms table name, function table_name()
would return hbase or kudu native table name.
Manually tested on mini-cluster with:
1. hdfs and s3 table with file format text and parquet.
2. hbase table.
3. kudu table.
Change-Id: If5da1112bcf38ae55b89eccfd7c7fad860819a99
Reviewed-on: http://gerrit.cloudera.org:8080/14660
Reviewed-by: Impala Public Jenkins <impala-public-jenk...@cloudera.com>
Tested-by: Impala Public Jenkins <impala-public-jenk...@cloudera.com>
> Scan node in HDFS profile should include name of table being scanned
> --------------------------------------------------------------------
>
> Key: IMPALA-9090
> URL: https://issues.apache.org/jira/browse/IMPALA-9090
> Project: IMPALA
> Issue Type: Improvement
> Components: Backend
> Reporter: Tim Armstrong
> Assignee: Xiaomeng Zhang
> Priority: Critical
>
> The only way to figure out the table being scanned by a scan node in the
> profile is to pull the string out of the explain plan or execsummary. This is
> awkward, both for manual and automated analysis of the profiles. We should
> include the table name as a string in the SCAN_NODE implementation.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-all-unsubscr...@impala.apache.org
For additional commands, e-mail: issues-all-h...@impala.apache.org
