[
https://issues.apache.org/jira/browse/HIVE-11388?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15066964#comment-15066964
]
Alan Gates commented on HIVE-11388:
-----------------------------------
Ok, that makes sense. Since the row will never really appear in the database
then the hostname field may not make sense. A couple of other thoughts:
# This "lock" will be limited by any intervening commits. This is probably ok
as we probably shouldn't be locking across commits anyway, but we want to make
sure we aren't inadvertently dropping the lock midway through.
# Can we make this a couple of protected methods in TxnHandler {{void
getMutex(String mutexName)}} and {{void releaseMutex(String mutexName)}}. This
way when the HBaseMetastore work gets around to implementing TxnHandler and has
to do it in a completely different way it will still have internal method to do
it with. If this adds a bunch of work you can ignore the request.
> there should only be 1 Initiator for compactions per Hive installation
> ----------------------------------------------------------------------
>
> Key: HIVE-11388
> URL: https://issues.apache.org/jira/browse/HIVE-11388
> Project: Hive
> Issue Type: Bug
> Components: Transactions
> Affects Versions: 1.0.0
> Reporter: Eugene Koifman
> Assignee: Eugene Koifman
> Priority: Critical
>
> org.apache.hadoop.hive.ql.txn.compactor.Initiator is a thread that runs
> inside the metastore service to manage compactions of ACID tables. There
> should be exactly 1 instance of this thread (even with multiple Thrift
> services).
> This is documented in
> https://cwiki.apache.org/confluence/display/Hive/Hive+Transactions#HiveTransactions-Configuration
> but not enforced.
> Should add enforcement, since more than 1 Initiator could cause concurrent
> attempts to compact the same table/partition - which will not work.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
