[
https://issues.apache.org/jira/browse/HBASE-2670?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Todd Lipcon updated HBASE-2670:
-------------------------------
Attachment: hbase-2670.txt
It turns out that the majority of this issue was that the comparator used by
MemStore didn't take into account the "memstore logical timestamp"
(memstoreTS). Thus, a new writer could overwrite older entries if it did a
write in the same millisecond. A concurrent reader would then see a partial row
because the new entries would be invisible, thus "revealing" cells with a lower
"real" TS.
This patch adds two tests to TestMemStore that check for the correct behavior.
I believe there is still a separate issue with multi-row scans and
updateReaders(), but I'll open a separate JIRA with separate tests for that one.
> Reader atomicity broken in trunk and branch
> -------------------------------------------
>
> Key: HBASE-2670
> URL: https://issues.apache.org/jira/browse/HBASE-2670
> Project: HBase
> Issue Type: Bug
> Components: regionserver
> Affects Versions: 0.20.5, 0.21.0
> Reporter: Todd Lipcon
> Assignee: Todd Lipcon
> Priority: Blocker
> Attachments: hbase-2670.txt
>
>
> 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.
