hbase-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Vladimir Rodionov (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HBASE-13890) Get/Scan from MemStore only (Client API)
Date Tue, 07 Jul 2015 16:38:05 GMT

    [ https://issues.apache.org/jira/browse/HBASE-13890?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14616950#comment-14616950

Vladimir Rodionov commented on HBASE-13890:

I see.. Checking the patch now. So this will fail it to client.. Can the get op be repeated
(with out memstore only setting) at server side only?
Yes, I think it can be improved. I am working on patch #2. I want to clarify little bit what
is this patch is for.

This is mostly to improve high performance counters (HPC), not Get, not Append (is anybody
using them anyway) and not Scan operations. Most recent version of HPCs are always in Memstore
(99.99% of a time), but each store file in this region/cf has its version as well (before
major compaction). When HBase reads counter it has to go through all store files and compare
results - very inefficient. This patch allows to bypass store files completely most of the


> Get/Scan from MemStore only (Client API)
> ----------------------------------------
>                 Key: HBASE-13890
>                 URL: https://issues.apache.org/jira/browse/HBASE-13890
>             Project: HBase
>          Issue Type: New Feature
>          Components: API, Client, Scanners
>            Reporter: Vladimir Rodionov
>            Assignee: Vladimir Rodionov
>         Attachments: HBASE-13890-v1.patch
> This is short-circuit read for get/scan when recent data (version) of a cell can be found
only in MemStore (with very high probability). 
> Good examples are: Atomic counters and appends. This feature will allow to bypass completely
store file scanners and improve performance and latency.

This message was sent by Atlassian JIRA

View raw message