Mailing-List: contact reviews-help@impala.incubator.apache.org; run by ezmlm
Precedence: bulk
Message-Id: <201706160550.v5G5otQO020767@ip-10-146-233-104.ec2.internal>
Date: Fri, 16 Jun 2017 05:50:55 +0000
From: "Alex Behm (Code Review)" <gerrit@cloudera.org>
To: Thomas Tauber-Marshall <tmarshall@cloudera.com>, impala-cr@cloudera.com,
        reviews@impala.incubator.apache.org
CC: Matthew Jacobs <mj@cloudera.com>
Reply-To: alex.behm@cloudera.com
Subject: =?UTF-8?Q?=5BImpala-ASF-CR=5D_IMPALA-4622=3A_Add_ALTER_COLUMN_statement=2E=0A?=
In-Reply-To: <gerrit.1495487248000.Id2e8bd65342b79644a0fdcd925e6f17797e89ad6@gerrit.cloudera.org>
References: <gerrit.1495487248000.Id2e8bd65342b79644a0fdcd925e6f17797e89ad6@gerrit.cloudera.org>
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Content-Disposition: inline
User-Agent: Gerrit/2.12.7
archived-at: Fri, 16 Jun 2017 05:50:59 -0000

Alex Behm has posted comments on this change.

Change subject: IMPALA-4622: Add ALTER COLUMN statement.
......................................................................


Patch Set 5:

(6 comments)

http://gerrit.cloudera.org:8080/#/c/6955/5/fe/src/main/java/org/apache/impala/analysis/AlterTableAlterColStmt.java
File fe/src/main/java/org/apache/impala/analysis/AlterTableAlterColStmt.java:

Line 48: 
remove blank line


Line 113:     for (Column c : t.getClusteringColumns()) {
Instead of this you can pass the Column into Table.isClusteringColumn() on L125


http://gerrit.cloudera.org:8080/#/c/6955/4/fe/src/main/java/org/apache/impala/analysis/AlterTableChangeColStmt.java
File fe/src/main/java/org/apache/impala/analysis/AlterTableChangeColStmt.java:

Line 56
> The behavior is the same either way - if the column is nullable, it gets th
Interesting. Thanks for explaining.


http://gerrit.cloudera.org:8080/#/c/6955/5/testdata/workloads/functional-query/queries/QueryTest/kudu_alter.test
File testdata/workloads/functional-query/queries/QueryTest/kudu_alter.test:

Line 497: # set a non-nullable column's default to null
also explain that this is equivalent do DROP DEFAULT which is consistent with Postgres behavior


http://gerrit.cloudera.org:8080/#/c/6955/5/tests/query_test/test_kudu.py
File tests/query_test/test_kudu.py:

Line 366:       pytest.skip("only runs in exhaustive")
Only runs in exhaustive to reduce core time.


Line 388:         cursor.execute("""insert into %s values (0, true, 0, 0, 0, 0, 0, 0, '0',
Doesn't this result in a PK conflict? I suggest adding a new row in every iteration.


-- 
To view, visit http://gerrit.cloudera.org:8080/6955
To unsubscribe, visit http://gerrit.cloudera.org:8080/settings

Gerrit-MessageType: comment
Gerrit-Change-Id: Id2e8bd65342b79644a0fdcd925e6f17797e89ad6
Gerrit-PatchSet: 5
Gerrit-Project: Impala-ASF
Gerrit-Branch: master
Gerrit-Owner: Thomas Tauber-Marshall <tmarshall@cloudera.com>
Gerrit-Reviewer: Alex Behm <alex.behm@cloudera.com>
Gerrit-Reviewer: Matthew Jacobs <mj@cloudera.com>
Gerrit-Reviewer: Thomas Tauber-Marshall <tmarshall@cloudera.com>
Gerrit-HasComments: Yes