[
https://issues.apache.org/jira/browse/HBASE-2670?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12876789#action_12876789
]
ryan rawson commented on HBASE-2670:
------------------------------------
#1 is the old memstore scanner - it would get an 'entire row' at a time and
then chomp on it's internal array. This made memstore scans really slow. The
whole index hbase saga.
#2 workish, and is what is implemented in HBASE-2616 (which is now in branch
and trunk).
The problem with #2, and you will see it in HRegionScanner is we no longer
update the read point "between rows". Since by the time we figure out we are
on the next row, we've already peek()ed the value off the scanner, we cant
switch to a different read point, or else you will get the first KeyValue from
one read point, and the rest from another (sounds familiar?)
As for why did I commit and not await review, this is because the hudson
instance at hudson.hbase.org is one of the few machines that trips against this
failure frequently. I haven't been able to repro on my dev environment, ever.
And now we have 11 successful branch builds of 0.20, thus indicating that our
bug is now fixed.
> Reader atomicity broken in trunk
> --------------------------------
>
> Key: HBASE-2670
> URL: https://issues.apache.org/jira/browse/HBASE-2670
> Project: HBase
> Issue Type: Bug
> Components: regionserver
> Affects Versions: 0.21.0
> Reporter: Todd Lipcon
> Assignee: Todd Lipcon
> Priority: Blocker
>
> There appears to be a bug in HBASE-2248 as committed to trunk. See following
> failing test:
> http://hudson.zones.apache.org/hudson/job/HBase-TRUNK/1296/testReport/junit/org.apache.hadoop.hbase/TestAcidGuarantees/testAtomicity/
> Think this is the same bug we saw early on in 2248 in the 0.20 branch, looks
> like the fix didn't make it over.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
