[
https://issues.apache.org/jira/browse/HIVE-18885?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16414129#comment-16414129
]
Vihang Karajgaonkar commented on HIVE-18885:
--------------------------------------------
Created a separate JIRA for cleaner thread improvements. Will update this patch
to only remove the java lock and update the comment of this class.
It is interesting to read the outdated java doc of this class. The original
DbNotificationListener seems to have asynchronous way of generating the
notifications which don't block the transactions. I guess somewhere down the
line this was repurposed for transactional notification events. I am curious to
look at the older version of this listener to see if there are certain things
which we can incorporate
> DbNotificationListener has a deadlock between Java and DB locks (2.x line)
> --------------------------------------------------------------------------
>
> Key: HIVE-18885
> URL: https://issues.apache.org/jira/browse/HIVE-18885
> Project: Hive
> Issue Type: Bug
> Components: Hive, Metastore
> Affects Versions: 2.3.2
> Reporter: Alexander Kolbasov
> Assignee: Vihang Karajgaonkar
> Priority: Major
> Attachments: HIVE-18885.01.branch-2.patch,
> HIVE-18885.02.branch-2.patch
>
>
> You can see the problem from looking at the code, but it actually created
> severe problems for real life Hive user.
> When {{alter table}} has {{cascade}} option it does the following:
> {code:java}
> msdb.openTransaction()
> ...
> List<Partition> parts = msdb.getPartitions(dbname, name, -1);
> for (Partition part : parts) {
> List<FieldSchema> oldCols = part.getSd().getCols();
> part.getSd().setCols(newt.getSd().getCols());
> String oldPartName =
> Warehouse.makePartName(oldt.getPartitionKeys(), part.getValues());
> updatePartColumnStatsForAlterColumns(msdb, part, oldPartName,
> part.getValues(), oldCols, part);
> msdb.alterPartition(dbname, name, part.getValues(), part);
> }
> {code}
> So it walks all partitions (and this may be huge list) and does some
> non-trivial operations in one single uber-transaction.
> When DbNotificationListener is enabled, it adds an event for each partition,
> all while
> holding a row lock on NOTIFICATION_SEQUENCE table. As a result, while this is
> happening no other write DDL can proceed. This can sometimes cause DB lock
> timeouts which cause HMS level operation retries which make things even worse.
> In one particular case this pretty much made HMS unusable.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
