hbase-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Mikhail Antonov (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HBASE-15648) Reduce number of concurrent region location lookups when MetaCache entry is cleared
Date Tue, 19 Apr 2016 17:07:25 GMT

    [ https://issues.apache.org/jira/browse/HBASE-15648?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15248194#comment-15248194
] 

Mikhail Antonov commented on HBASE-15648:
-----------------------------------------

If things are in meta cache, that will be handled at the top of #locateRegionInMeta()

Goes like this:

  - first try to get cached location; if found, return
  - grab lock for the regionName
  - try cached location again in case someone who held the lock retrieved it for us
  - if still no luck, do a meta scan

> Reduce number of concurrent region location lookups when MetaCache entry is cleared
> -----------------------------------------------------------------------------------
>
>                 Key: HBASE-15648
>                 URL: https://issues.apache.org/jira/browse/HBASE-15648
>             Project: HBase
>          Issue Type: Sub-task
>          Components: Client
>    Affects Versions: 1.3.0
>            Reporter: Mikhail Antonov
>            Assignee: Mikhail Antonov
>             Fix For: 1.3.0
>
>         Attachments: HBASE-15648-branch-1.3.v1.patch
>
>
> It seems in HConnectionImplementation#locateRegionInMeta if region location is removed
from the cache, with large number of client threads we could have many of them getting cache
miss and doing meta scan, which looks unnecessary - we could empty mechanism similar to what
we have in IdLock in HFileReader to fetch the block to cache, do ensure that if one thread
is already looking up location for region R1, other threads who need it's location wait until
first thread finishes his work.



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

Mime
View raw message