You are viewing a plain text version of this content. The canonical link for it is here.
Posted to notifications@asterixdb.apache.org by "Young-Seok Kim (JIRA)" <ji...@apache.org> on 2015/12/08 20:33:11 UTC
[jira] [Created] (ASTERIXDB-1218) Incoming ConcurrentLockManager's
OutOfMemoryError issue with index-only-plan query
Young-Seok Kim created ASTERIXDB-1218:
-----------------------------------------
Summary: Incoming ConcurrentLockManager's OutOfMemoryError issue with index-only-plan query
Key: ASTERIXDB-1218
URL: https://issues.apache.org/jira/browse/ASTERIXDB-1218
Project: Apache AsterixDB
Issue Type: Bug
Reporter: Young-Seok Kim
Assignee: Till Westmann
Currently, when there are so many lock requests from a single (index-only-plan) query, ConcurrentLockManager will encounter java.lang.OutOfMemoryError. This does happen due to the fact that index-only-plan query keeps all locks and release them when the query is completed. This doesn't happen for no-index-only-plan queries which acquire instantLock.
We should consider having lock escalation mechanism which escalates entity-level S-lock into dataset-level S-lock according to a threshold value or a policy in order to reduce the number of resources to be allocated by ConcurrentLockManager in such a situation.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)