[
https://issues.apache.org/jira/browse/HBASE-13297?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14370787#comment-14370787
]
Lars Hofhansl commented on HBASE-13297:
---------------------------------------
With this patch the client not longer calculates the size. As long as
_anything_ was returned from the server, we'll use that to fill cache and
return to the client. As a consequence if the scanner caching is set such that
we hit the size limit on the server first, we can longer tell on the client
that we exhausted the region by looking the result size, instead we need to
another RPC, which will return an empty result, and hence in the next iteration
nothing is returned from the server, which has us move to the next region.
Performance wise this should be acceptable, we'd get one extra RPC per region
crossing. I do not think that is avoidable.
> 0.98 and 1.0: Remove client side result size calculation
> --------------------------------------------------------
>
> Key: HBASE-13297
> URL: https://issues.apache.org/jira/browse/HBASE-13297
> Project: HBase
> Issue Type: Sub-task
> Components: Client
> Reporter: Lars Hofhansl
> Assignee: Lars Hofhansl
> Fix For: 1.0.1, 0.98.13
>
> Attachments: 13297-0.98.txt
>
>
> As described in parent, this can lead to missed rows when the client and
> server calculate different size values.
> The patch here proposes a backwards compatible patch for 0.98 and 1.0.x.
> Parent will do a patch for 1.1 and 2.0.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
