[
https://issues.apache.org/jira/browse/HBASE-10447?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
ramkrishna.s.vasudevan updated HBASE-10447:
-------------------------------------------
Attachment: HBASE-10447_trunk_1.patch
This should be the simplest patches. The constructor that is changed in this
patch is the one that is called over in Compaction and flush. In 0.96 case it
does not add the observers. So should be the case in 0.98 and trunk. Do we
need the reset to happen in compaction? I don't think so.
@apurtell,[~lhofhansl], [~anoop.hbase]
Pls review this patch and if we are ok we could commit this.
> Memstore flusher scans storefiles also when the scanner heap gets reset
> -----------------------------------------------------------------------
>
> Key: HBASE-10447
> URL: https://issues.apache.org/jira/browse/HBASE-10447
> Project: HBase
> Issue Type: Bug
> Affects Versions: 0.98.0, 0.99.0
> Reporter: ramkrishna.s.vasudevan
> Assignee: ramkrishna.s.vasudevan
> Priority: Blocker
> Fix For: 0.98.0, 0.99.0
>
> Attachments: HBASE-10447_0.98.patch, HBASE-10447_trunk.patch,
> HBASE-10447_trunk_1.patch
>
>
> See the mail thread
> http://osdir.com/ml/general/2014-01/msg61294.html
> In case of flush we create a memstore flusher which in turn creates a
> StoreScanner backed by a Single ton MemstoreScanner.
> But this scanner also registers for any updates in the reader in the HStore.
> Is this needed?
> If this happens then any update on the reader may nullify the current heap
> and the entire Scanner Stack is reset, but this time with the other scanners
> for all the files that satisfies the last top key. So the flush that happens
> on the memstore holds the storefile scanners also in the heap that was
> recreated but originally the intention was to create a scanner on the
> memstore alone.
--
This message was sent by Atlassian JIRA
(v6.1.5#6160)
