hbase-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Dave Latham (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HBASE-9931) Optional setBatch for CopyTable to copy large rows in batches
Date Thu, 21 Nov 2013 17:15:36 GMT

    [ https://issues.apache.org/jira/browse/HBASE-9931?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13829104#comment-13829104
] 

Dave Latham commented on HBASE-9931:
------------------------------------

I think a new config option to call Scan.setBatch would do it.  I'm always a bit puzzled about
how scanner batching and caching interact with mixed length rows, but I imagine it would work
out ok.  (4 years later, and still using a version where HBASE-1996 didn't make it.)

> Optional setBatch for CopyTable to copy large rows in batches
> -------------------------------------------------------------
>
>                 Key: HBASE-9931
>                 URL: https://issues.apache.org/jira/browse/HBASE-9931
>             Project: HBase
>          Issue Type: Improvement
>          Components: mapreduce
>            Reporter: Dave Latham
>
> We've had CopyTable jobs fail because a small number of rows are wide enough to not fit
into memory.  If we could specify the batch size for CopyTable scans that shoud be able to
break those large rows up into multiple iterations to save the heap.



--
This message was sent by Atlassian JIRA
(v6.1#6144)

Mime
View raw message