You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@ignite.apache.org by "Alexey Goncharuk (JIRA)" <ji...@apache.org> on 2017/06/23 09:24:00 UTC

[jira] [Created] (IGNITE-5579) Make sure identical binary metadata updates from the same node do not happen twice

Alexey Goncharuk created IGNITE-5579:
----------------------------------------

             Summary: Make sure identical binary metadata updates from the same node do not happen twice
                 Key: IGNITE-5579
                 URL: https://issues.apache.org/jira/browse/IGNITE-5579
             Project: Ignite
          Issue Type: Improvement
          Components: cache
    Affects Versions: 1.7
            Reporter: Alexey Goncharuk
             Fix For: 2.1


It is possible that multiple user threads attempt identical binary metadata update concurrently. In this case, a node will just increase contention on the cache key, but once the lock is acquired there will be nothing changed. On large topologies, this may lead to nodes waiting for the metadata update for hours (!).
We should work out a way to identify identical metadata updates and allow only one thread to proceed.



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)