[
https://issues.apache.org/jira/browse/HBASE-5974?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13287949#comment-13287949
]
Zhihong Yu commented on HBASE-5974:
-----------------------------------
HRegionInterface.java doesn't exist in trunk so patch v2 wouldn't apply to
trunk.
I would suggest creating patch for trunk and run through hadoop QA.
{code}
+ LOG.info("Seq number based scan API not present at RS side!
Trying with API: "
{code}
I think the above log should be at warn level.
{code}
+ } else if (ioe instanceof CallSequenceOutOfOrderException) {
+ // The callSeq from the client not matched with the one expected at
the RS side
+ // This means the RS might have done extra scanning of data which is
not received by the
+ // client.Throw a DNRE so that we close the current scanner and
opens a new one with RS.
+ throw new DoNotRetryIOException("Reset scanner", ioe);
{code}
Should we disclose a little more detail in the message of DNRIOE ? The above is
the same as response to NotServingRegionException and
RegionServerStoppedException.
'not matched with' -> 'does not match'
'is not received' -> 'has not been received'
'opens a new' -> 'open a new'
{code}
+ // if callSeq do not match throw Exception straight away. This needs to be
performed even
{code}
'do not match' -> 'does not match'
{code}
+public class TestClientScannerRPCTimesout {^M
{code}
Please add short javadoc for the test class. I think it should be called
TestClientScannerRPCTimeout.
Please use utility such as dos2unix to remove the trailing ^M from the patch
file.
{code}
+ public static class RegionServerWithScanTimesout extends
MiniHBaseClusterRegionServer {^M
{code}
The above class can be made private. It should be named
RegionServerWithScanTimeout.
{code}
+ * Thrown by a region server while scan related next() calls. Both client and
server maintain a^M
+ * callSequence and if the both do not match, RS will throw this exception.^M
+ */^M
+public class CallSequenceOutOfOrderException extends IOException {^M
{code}
CallSequenceOutOfOrderException should extend DoNotRetryIOException so that we
don't need to create DoNotRetryIOException instance (shown above).
'while scan related next()' -> 'while doing scan related next()'
'the both do not' -> 'they do not'
It would be nice for Todd to take a look at the patch.
> Scanner retry behavior with RPC timeout on next() seems incorrect
> -----------------------------------------------------------------
>
> Key: HBASE-5974
> URL: https://issues.apache.org/jira/browse/HBASE-5974
> Project: HBase
> Issue Type: Bug
> Components: client, regionserver
> Affects Versions: 0.90.7, 0.92.1, 0.94.0, 0.96.0
> Reporter: Todd Lipcon
> Assignee: Anoop Sam John
> Priority: Critical
> Fix For: 0.94.1
>
> Attachments: HBASE-5974_0.94.patch, HBASE-5974_94-V2.patch
>
>
> I'm seeing the following behavior:
> - set RPC timeout to a short value
> - call next() for some batch of rows, big enough so the client times out
> before the result is returned
> - the HConnectionManager stuff will retry the next() call to the same server.
> At this point, one of two things can happen: 1) the previous next() call will
> still be processing, in which case you get a LeaseException, because it was
> removed from the map during the processing, or 2) the next() call will
> succeed but skip the prior batch of rows.
--
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
