[
https://issues.apache.org/jira/browse/HBASE-8701?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13692642#comment-13692642
]
Jeffrey Zhong commented on HBASE-8701:
--------------------------------------
Thanks [[email protected]] for the detailed reviews!
{quote}
I see that the 'original seqnum' IS serialized, written into WALKey.
{quote}
Yeah. That's introduced by the latest patch to handle chained failure
scenarios. That's why you see me the "original sequence number" passed down by
WALEdit. If we use "skipWal" for edits replay, we won't need to do this and
also might achieve even better performance. Let me know which way you lean
towards.(I personally prefer skipwal option)
{quote}
Seems like any non-null mvcc needs 'decoding'. If so, do we need this flag
stuff going on?
{quote}
Not every one because most case mvcc has values 0 unless we have recovery
edits. mvcc optimization normally writes mvcc out as 0. Since we pass negative
mvcc, the MAX_MEMSTORE_TS_KEY will be 0. In that case, hfilereader won't read
the negative mvcc value at all which against the purpose to use mvcc to store
negative sequence number. I'll change the boolean decodeMemstoreTS to a long
MIN_MEMSTORE_TS_KEY to be more clear on whether a hfile contains negative mvcc
values.
{quote}
I have already remarked on how it is odd that HRegion adds itself to the
recovering region list – that the regionserver should be doing this, not the
hregion (it used to be done in OpenRegionHandler?)
{quote}
I'll modify this part.
{quote}
Previous we'd get the seqid internal from the left and right key but not we
allow the seqid being passed independent of the kvs being compared. Is that
what we want?
{quote}
This is to pass enough information to the compare() method in calls where only
KeyValue's were passed before.
{quote}
MINOR_VERSION_WITH_MVCC_SEQ_ID_UNION is defined but not used?
{quote}
It serves the purpose for developers to know what purpose minor version 4
serves. Having a constant for version 4 is in line with prior practice.
{quote}
This means that the seqid we flush with is never -ve?
{quote}
That's correct. I can add an assertion here.
I'll try to provide a new patch based on your feedbacks. Thanks.
> distributedLogReplay need to apply wal edits in the receiving order of those
> edits
> ----------------------------------------------------------------------------------
>
> Key: HBASE-8701
> URL: https://issues.apache.org/jira/browse/HBASE-8701
> Project: HBase
> Issue Type: Bug
> Components: MTTR
> Reporter: Jeffrey Zhong
> Assignee: Jeffrey Zhong
> Fix For: 0.98.0, 0.95.2
>
> Attachments: 8701-v3.txt, hbase-8701-v4.patch, hbase-8701-v5.patch,
> hbase-8701-v6.patch, hbase-8701-v7.patch
>
>
> This issue happens in distributedLogReplay mode when recovering multiple puts
> of the same key + version(timestamp). After replay, the value is
> nondeterministic of the key
> h5. The original concern situation raised from [~eclark]:
> For all edits the rowkey is the same.
> There's a log with: [ A (ts = 0), B (ts = 0) ]
> Replay the first half of the log.
> A user puts in C (ts = 0)
> Memstore has to flush
> A new Hfile will be created with [ C, A ] and MaxSequenceId = C's seqid.
> Replay the rest of the Log.
> Flush
> The issue will happen in similar situation like Put(key, t=T) in WAL1 and
> Put(key,t=T) in WAL2
> h5. Below is the option(proposed by Ted) I'd like to use:
> a) During replay, we pass original wal sequence number of each edit to the
> receiving RS
> b) In receiving RS, we store negative original sequence number of wal edits
> into mvcc field of KVs of wal edits
> c) Add handling of negative MVCC in KVScannerComparator and KVComparator
> d) In receiving RS, write original sequence number into an optional field of
> wal file for chained RS failure situation
> e) When opening a region, we add a safety bumper(a large number) in order for
> the new sequence number of a newly opened region not to collide with old
> sequence numbers.
> In the future, when we stores sequence number along with KVs, we can adjust
> the above solution a little bit by avoiding to overload MVCC field.
> h5. The other alternative options are listed below for references:
> Option one
> a) disallow writes during recovery
> b) during replay, we pass original wal sequence ids
> c) hold flush till all wals of a recovering region are replayed. Memstore
> should hold because we only recover unflushed wal edits. For edits with same
> key + version, whichever with larger sequence Id wins.
> Option two
> a) During replay, we pass original wal sequence ids
> b) for each wal edit, we store each edit's original sequence id along with
> its key.
> c) during scanning, we use the original sequence id if it's present otherwise
> its store file sequence Id
> d) compaction can just leave put with max sequence id
> Please let me know if you have better ideas.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
