[
https://issues.apache.org/jira/browse/HBASE-2214?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13423351#comment-13423351
]
Jean-Daniel Cryans commented on HBASE-2214:
-------------------------------------------
Following this discussion on user@
http://permalink.gmane.org/gmane.comp.java.hadoop.hbase.user/27936
If the max scanner result size is only set on the region server it's possible
miss rows. Even with the latest patch in this jira it's still possible since
the client doesn't tell the server it's going to use the default instead of
whatever the server is using.
Maybe the {{ClientScanner}} could modify the {{Scan}} that it passes the max
scanner result it uses to the region server
> Do HBASE-1996 -- setting size to return in scan rather than count of rows --
> properly
> -------------------------------------------------------------------------------------
>
> Key: HBASE-2214
> URL: https://issues.apache.org/jira/browse/HBASE-2214
> Project: HBase
> Issue Type: New Feature
> Reporter: stack
> Fix For: 0.96.0, 0.94.2
>
> Attachments: HBASE-2214-0.94-v2.txt, HBASE-2214-0.94-v3.txt,
> HBASE-2214-0.94.txt, HBASE-2214-v4.txt, HBASE-2214-v5.txt, HBASE-2214-v6.txt,
> HBASE-2214-v7.txt, HBASE-2214_with_broken_TestShell.txt
>
>
> The notion that you set size rather than row count specifying how many rows a
> scanner should return in each cycle was raised over in hbase-1966. Its a
> good one making hbase "regular" though the data under it may vary.
> HBase-1966 was committed but the patch was constrained by the fact that it
> needed to not change RPC interface. This issue is about doing hbase-1966 for
> 0.21 in a clean, unconstrained way.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators:
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
