hbase-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Lars Hofhansl (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HBASE-6371) [89-fb] Tier based compaction
Date Thu, 18 Oct 2012 00:48:03 GMT

    [ https://issues.apache.org/jira/browse/HBASE-6371?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13478537#comment-13478537
] 

Lars Hofhansl commented on HBASE-6371:
--------------------------------------

Is this in the 0.89 branch, and is it finished?

Does this only change the compaction selection?

A feature I'd be interested in is separating old and new data, such it would not typically
be compacted into the same file(s) - even in a major compaction.
If there is hot new data and lot of older data that is only occasionally queried, if old and
new data ends up in the same set of files, scanning is slow, because scanning the latest version
only is hitting many more blocks then necessary and old versions of KVs need to be skipped
(and HFiles with only old data could otherwise simply be ignored).

An approach LevelDB might be interesting. Here's a description: http://code.google.com/p/leveldb/source/browse/doc/impl.html

Is that what we are trying to do with this?
                
> [89-fb] Tier based compaction
> -----------------------------
>
>                 Key: HBASE-6371
>                 URL: https://issues.apache.org/jira/browse/HBASE-6371
>             Project: HBase
>          Issue Type: Improvement
>            Reporter: Akashnil
>            Assignee: Liyin Tang
>              Labels: noob
>
> Currently, the compaction selection is not very flexible and is not sensitive to the
hotness of the data. Very old data is likely to be accessed less, and very recent data is
likely to be in the block cache. Both of these considerations make it inefficient to compact
these files as aggressively as other files. In some use-cases, the access-pattern is particularly
obvious even though there is no way to control the compaction algorithm in those cases.
> In the new compaction selection algorithm, we plan to divide the candidate files into
different levels according to oldness of the data that is present in those files. For each
level, parameters like compaction ratio, minimum number of store-files in each compaction
may be different. Number of levels, time-ranges, and parameters for each level will be configurable
online on a per-column family basis.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

Mime
View raw message