[
https://issues.apache.org/jira/browse/HIVE-20006?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16541994#comment-16541994
]
Ashutosh Chauhan commented on HIVE-20006:
-----------------------------------------
[~jcamachorodriguez] Does the description still accurately reflects the changes
in the patch? If not, can you please update it.
> Make materializations invalidation cache work with multiple active remote
> metastores
> ------------------------------------------------------------------------------------
>
> Key: HIVE-20006
> URL: https://issues.apache.org/jira/browse/HIVE-20006
> Project: Hive
> Issue Type: Improvement
> Components: Materialized views
> Affects Versions: 3.0.0
> Reporter: Jesus Camacho Rodriguez
> Assignee: Jesus Camacho Rodriguez
> Priority: Critical
> Attachments: HIVE-19027.01.patch, HIVE-19027.02.patch,
> HIVE-19027.03.patch, HIVE-19027.04.patch, HIVE-20006.01.patch,
> HIVE-20006.02.patch, HIVE-20006.03.patch, HIVE-20006.04.patch,
> HIVE-20006.05.patch, HIVE-20006.06.patch, HIVE-20006.patch
>
>
> The main points:
> - Only MVs stored in transactional tables can have a time window value of 0.
> Those are the only MVs that can be guaranteed to not be outdated when a query
> is executed, if we use custom storage handlers to store the materialized
> view, we cannot make any promises.
> - For MVs that +cannot be outdated+, we do not check the metastore. Instead,
> comparison is based on valid write id lists.
> - For MVs that +can be outdated+, we still rely on the invalidation cache.
> ** The window for valid outdated MVs can be specified in intervals of 1
> minute (less than that, it is difficult to have any guarantees about whether
> the MV is actually outdated by less than a minute or not).
> ** The async loading is done every interval / 2 (or probably better, we can
> make it configurable).
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
