cassandra-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Paulo Motta (JIRA)" <j...@apache.org>
Subject [jira] [Updated] (CASSANDRA-14251) View replica is not written to pending endpoint when base replica is also view replica
Date Wed, 21 Feb 2018 23:22:00 GMT

     [ https://issues.apache.org/jira/browse/CASSANDRA-14251?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]

Paulo Motta updated CASSANDRA-14251:
------------------------------------
    Reproduced In: 3.11.1, 3.0.15  (was: 3.0.15, 3.11.1)
           Status: Patch Available  (was: Open)

Added regression [dtest|https://github.com/pauloricardomg/cassandra-dtest/commit/dbf6eba59da67edaa50930046fc440430fe61d2d]
and [restored|https://github.com/pauloricardomg/cassandra/commit/1a06c97a7896621f346b806e6369a820b75fdd75]
correct behavior along with a NEWS.txt asking MV users potentially affected by this to run
repair on the view to ensure all replicas are up to date.

||3.0||3.11||trunk||dtest||
|[branch|https://github.com/apache/cassandra/compare/cassandra-3.0...pauloricardomg:3.0-14251]|[branch|https://github.com/apache/cassandra/compare/cassandra-3.11...pauloricardomg:3.11-14251]|[branch|https://github.com/apache/cassandra/compare/trunk...pauloricardomg:trunk-14251]|[branch|https://github.com/apache/cassandra-dtest/compare/master...pauloricardomg:14251]|

Will run internal CI and post results here when ready.

> View replica is not written to pending endpoint when base replica is also view replica
> --------------------------------------------------------------------------------------
>
>                 Key: CASSANDRA-14251
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-14251
>             Project: Cassandra
>          Issue Type: Bug
>          Components: Materialized Views
>            Reporter: Paulo Motta
>            Assignee: Paulo Motta
>            Priority: Major
>             Fix For: 4.0, 3.0.17, 3.11.3
>
>
> From the [dev list|https://www.mail-archive.com/dev@cassandra.apache.org/msg12084.html]:
> bq. There's an optimization that when we're lucky enough that the paired view replica
is the same as this base replica, mutateMV doesn't use the normal view-mutation-sending code
(wrapViewBatchResponseHandler) and just writes the mutation locally. In particular, in this
case we do NOT write to the pending node (unless I'm missing something). But, sometimes all
replicas will be paired with themselves - this can happen for example when number of nodes
is equal to RF, or when the base and view table have the same partition keys (but different
clustering keys). In this case, it seems the pending node will not be written at all...
> This was a regression from CASSANDRA-13069 and the original behavior should be restored.



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@cassandra.apache.org
For additional commands, e-mail: commits-help@cassandra.apache.org


Mime
View raw message