[jira] [Commented] (KUDU-38) bootstrap should not replay logs that are known to be fully flushed

["Todd Lipcon (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/KUDU-38?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15419421#comment-15419421 ] 

Todd Lipcon commented on KUDU-38:
---------------------------------

Thinking about this a bit more... I think the best bet is:
- when flushing TabletMetadata, include the lowest live log anchor required for durability (which is often newer than the oldest log segment being retained due to time-based or segment-count-based retention)
- when rolling log segments, issue an msync on the log *index* file, so that any complete segments are known to be fully indexed.
- on restart, only start replaying from the log segment that contains that op index
-- we still need to copy over the older logs, perhaps via hard-linking them into the recovery directory

> bootstrap should not replay logs that are known to be fully flushed
> -------------------------------------------------------------------
>
>                 Key: KUDU-38
>                 URL: https://issues.apache.org/jira/browse/KUDU-38
>             Project: Kudu
>          Issue Type: Improvement
>          Components: tablet
>    Affects Versions: M3
>            Reporter: Todd Lipcon
>            Assignee: Todd Lipcon
>
> Currently the bootstrap process will process all of the log segments, including those that can be trivially determined to contain only durable edits. This makes startup unnecessarily slow.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)