[
https://issues.apache.org/jira/browse/HBASE-17584?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Duo Zhang updated HBASE-17584:
------------------------------
Attachment: HBASE-17584.patch
Add a reset flag for ScanMetrics.getScanMetrics. Do not reset ScanMetrics when
publish it to the Scan object. So you can still Scan.getScanMetrics to get the
metrics. But if you use ResultScanner.getScanMetrics and reset the counters,
the metrics published to Scan object will be messed up. But I think this is
acceptable? If you use ResultScanner.getScanMetrics get the metrics, then you
do not need to use Scan.getScanMetrics anymore.
> Expose ScanMetrics with ResultScanner rather than Scan
> ------------------------------------------------------
>
> Key: HBASE-17584
> URL: https://issues.apache.org/jira/browse/HBASE-17584
> Project: HBase
> Issue Type: Sub-task
> Components: Client, mapreduce, scan
> Affects Versions: 2.0.0, 1.4.0
> Reporter: Duo Zhang
> Assignee: Duo Zhang
> Fix For: 2.0.0, 1.4.0
>
> Attachments: HBASE-17584.patch
>
>
> I think this have been discussed many times... It is a bad practice to
> directly modify the Scan object passed in when calling getScanner. The reason
> that we can not use a copy is we need to use the Scan object to expose scan
> metrics. So we need to find another way to expose the metrics.
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
