[
https://issues.apache.org/jira/browse/PHOENIX-916?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13962620#comment-13962620
]
Lars Hofhansl commented on PHOENIX-916:
---------------------------------------
This is generally an issue in HBase when the clock skew between RegionServers
is larger than the time it take to move a region between servers.
Is 1s enough? (this is in the same ballpark as the time it takes to move a
region, so not sure whether that is good default).
> Time clock skew issue when SYSTEM.CATALOG moves between region servers
> ----------------------------------------------------------------------
>
> Key: PHOENIX-916
> URL: https://issues.apache.org/jira/browse/PHOENIX-916
> Project: Phoenix
> Issue Type: Bug
> Affects Versions: 3.0.0, 4.0.0, 5.0.0
> Reporter: Jeffrey Zhong
> Assignee: Jeffrey Zhong
> Priority: Critical
> Attachments: phoenix-916.patch
>
>
> Since we rely on the sever time stamp of RS which is hosting SYSTEM.CATALOG.
> When the table moves to another region server, the server time may
> decrease(increasing is all right). This may cause a query doesn't return
> anything though it should be.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
