hive-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Szehon Ho" <sze...@cloudera.com>
Subject Re: Review Request 23820: HIVE-7445:Improve LOGS for Hive when a query is not able to acquire locks
Date Tue, 22 Jul 2014 22:33:15 GMT

-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/23820/#review48441
-----------------------------------------------------------


Hey Chaoyu thanks, the logic looks good now.  Can you also fix all the red-space?  It's two
lines per indent in hive, thanks.


ql/src/java/org/apache/hadoop/hive/ql/lockmgr/zookeeper/ZooKeeperHiveLockManager.java
<https://reviews.apache.org/r/23820/#comment85045>

    Why was this necessary to change?


- Szehon Ho


On July 22, 2014, 7:21 p.m., Chaoyu Tang wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/23820/
> -----------------------------------------------------------
> 
> (Updated July 22, 2014, 7:21 p.m.)
> 
> 
> Review request for hive, Brock Noland, Prasad Mujumdar, Szehon Ho, and Xuefu Zhang.
> 
> 
> Bugs: HIVE-7445
>     https://issues.apache.org/jira/browse/HIVE-7445
> 
> 
> Repository: hive-git
> 
> 
> Description
> -------
> 
> This patch enables ZookeeperHiveLockManager.java, when in debug mode, to log out information
about contentious locks if lockmgr fails to acquire a lock for a query. The changes include:
> 1. Collect and log out contentious lock information in ZookeeperHiveLockManager.java
> 2. Improve lock retry counting and existing logging, and modified some qtests output
files to match these logging
>  
> 
> 
> Diffs
> -----
> 
>   ql/src/java/org/apache/hadoop/hive/ql/lockmgr/HiveLockObject.java 1cc3074 
>   ql/src/java/org/apache/hadoop/hive/ql/lockmgr/zookeeper/ZooKeeperHiveLockManager.java
65b9136 
>   ql/src/test/results/clientnegative/insert_into1.q.out a38b679 
>   ql/src/test/results/clientnegative/insert_into2.q.out f21823a 
>   ql/src/test/results/clientnegative/insert_into3.q.out 5f1581e 
>   ql/src/test/results/clientnegative/insert_into4.q.out 5dcdd50 
>   ql/src/test/results/clientnegative/lockneg1.q.out 6a76cd7 
>   ql/src/test/results/clientnegative/lockneg_try_db_lock_conflict.q.out a9833a8 
>   ql/src/test/results/clientnegative/lockneg_try_drop_locked_db.q.out d67365a 
>   ql/src/test/results/clientnegative/lockneg_try_lock_db_in_use.q.out 89d3265 
> 
> Diff: https://reviews.apache.org/r/23820/diff/
> 
> 
> Testing
> -------
> 
> 1. Manual tests: following debug logging was printed out as expected during a lock acquisition
failure.
> ---
> Unable to acquire IMPLICIT, EXCLUSIVE lock default@sample_07 after 2 attempts.
> 14/07/22 11:44:40 ERROR ZooKeeperHiveLockManager: Unable to acquire IMPLICIT, EXCLUSIVE
lock default@sample_07 after 2 attempts.
> 14/07/22 11:44:40 DEBUG ZooKeeperHiveLockManager: Requested lock default@sample_07::
mode:IMPLICIT,EXCLUSIVE; query:insert into table sample_07 select * from sample_08
> 14/07/22 11:44:40 DEBUG ZooKeeperHiveLockManager: Conflicting lock to default@sample_07::
mode:IMPLICIT;query:select code from sample_07;queryId:root_20140722084242_98f8d9d7-d110-45c0-8c8b-12da2e5172d9;clientIp:10.20.92.233
> ---
> 
> 2. Precommit tests
> The test failure testCliDriver_ql_rewrite_gbtoidx is a preexisting one and I think it
is not related to this change.
> 
> 
> Thanks,
> 
> Chaoyu Tang
> 
>


Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message