[
https://issues.apache.org/jira/browse/HBASE-2283?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12844672#action_12844672
]
stack commented on HBASE-2283:
------------------------------
Here's some comments.
I think KeyValueList not the best name for a List that is purposed to appending
to a WAL log. KVL implies List<KeyValue> which it replaces in your code but in
fact it does much more. For one its a Writable. It also takes care of your
fancy new transaction start/end markings. Rename it WALEdit or HLogEdit or
something? Should it also be moved to the wal package?
(o.a.h.h.regionserver.wal).
Minor: Should it implement List<KeyValue>? Would that help? You wouldn't need
to have a getList?
I love the class comment on KVL.
readNonLengthData should instead be an overloading of readFields, just add the
length param?
Minor: The following if (kvlist.size() > 0) { is usually cheaper if done as
!kvlist.isEmpty IIRC, especially if the list one of the concurrent
implementations where size calc is expensive (probably not pertinent here).
Its interesting, our number-of-entries in hlog will change now, now you group
up puts, etc.
Patch looks great Kannan.
> row level atomicity
> --------------------
>
> Key: HBASE-2283
> URL: https://issues.apache.org/jira/browse/HBASE-2283
> Project: Hadoop HBase
> Issue Type: Bug
> Reporter: Kannan Muthukkaruppan
> Priority: Blocker
> Fix For: 0.20.4, 0.21.0
>
> Attachments: rowLevelAtomicity_2283_v1.patch
>
>
> The flow during a HRegionServer.put() seems to be the following. [For now,
> let's just consider single row Put containing edits to multiple column
> families/columns.]
> HRegionServer.put() does a:
> HRegion.put();
> syncWal() (the HDFS sync call). /* this is assuming we have HDFS-200
> */
> HRegion.put() does a:
> for each column family
> {
> HLog.append(all edits to the colum family);
> write all edits to Memstore;
> }
> HLog.append() does a :
> foreach edit in a single column family {
> doWrite()
> }
> doWrite() does a:
> this.writer.append().
> There seems to be two related issues here that could result in
> inconsistencies.
> Issue #1: A put() does a bunch of HLog.append() calls. These in turn do a
> bunch of "write" calls on the underlying DFS stream. If we crash after
> having written out some append's to DFS, recovery will run and apply a
> partial transaction to memstore.
> Issue #2: The updates to memstore should happen after the sync rather than
> before. Otherwise, there is the danger that the write to DFS (sync) fails for
> some reason & we return an error to the client, but we have already taken
> edits to the memstore. So subsequent reads will serve uncommitted data.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
