[ https://issues.apache.org/jira/browse/PHOENIX-3872?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16021934#comment-16021934 ]
Sergey Soldatov commented on PHOENIX-3872: ------------------------------------------ [~giacomotaylor] I don't have objections. I'm just curious when it may be required. The cache is used during preBatch, so if we get cache expired exception that means that mutations were in the queue for this period of time or longer (30 minutes atm?). Usually such kind of delays is a bad sign and may require an investigation why it happen. Not sure, but I have feeling that in this case the client will be timed out as well and would not clean the cache on close(). As for PHOENIX-3875, I believe we saw index metadata errors during splits and merges, but need to refresh my memory about it. > Increase default for index metadata expiration > ---------------------------------------------- > > Key: PHOENIX-3872 > URL: https://issues.apache.org/jira/browse/PHOENIX-3872 > Project: Phoenix > Issue Type: Bug > Reporter: James Taylor > Assignee: James Taylor > Fix For: 4.11.0 > > Attachments: PHOENIX-3872.patch > > > The default timeout is 30 minutes. We should increase it substantially to > ride out any retry loops that may occur on the server side. 2 hours would > probably be a good default. -- This message was sent by Atlassian JIRA (v6.3.15#6346)