You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@ignite.apache.org by "Vladimir Ozerov (JIRA)" <ji...@apache.org> on 2015/03/20 10:03:39 UTC

[jira] [Created] (IGNITE-523) Cache operation is stuck when an exception is thrown from continuous query filter.

Vladimir Ozerov created IGNITE-523:
--------------------------------------

             Summary: Cache operation is stuck when an exception is thrown from continuous query filter.
                 Key: IGNITE-523
                 URL: https://issues.apache.org/jira/browse/IGNITE-523
             Project: Ignite
          Issue Type: Bug
          Components: cache
    Affects Versions: sprint-1
            Reporter: Vladimir Ozerov
            Priority: Critical
             Fix For: sprint-2


Test scenario:
1) Start 2 nodes with ATOMIC cache and no backups.
2) Start continuous query with filter which throws an exception (say, CacheEntryListenerException as per contract).
3) Put near key into cache.
4) Observe that you got stuck.

Root cause: CacheContinuousQuerHandler invokes filter in register()/onEntryUpdate() method. Exception is thrown form here to GridDhtAtomicCache.
Atomic cache "updateSingle()" method, in turn, catches only IgniteCheckedException. As a result, filter exception is propagated to IO manager and simply ignored there. Aa a final result, update response is never sent to remote node.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)