[
https://issues.apache.org/jira/browse/HBASE-8521?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13742519#comment-13742519
]
Jean-Marc Spaggiari commented on HBASE-8521:
--------------------------------------------
Hi [~jdcryans], thanks for looking at it.
{quote}
There are 11 "== Durability.USE_DEFAULT", can we just have a method
somewhere (like in Mutation) that does it and is named "isDefaultDurability"?
{quote}
This is because setWriteToWAL is now deprecated. If I set something locally to
do the same thing as what setWriteToWAL was doing, then it's not very clean
since we are just bypassing the deprecation warning. Other option is to keep
setWriteToWAL calls and add @suppressWarning to avoid them and keep
setWriteToWAL?
{quote}
LoadIncrementalHFiles.assignSeqIds should be final.
{quote}
I don't think we can. If I put it final, how are you going to modify it on the
constructor? (assignSeqIds = conf.getBoolean(ASSIGN_SEQ_IDS, true))
{quote}
Since bulk loaded files can have sequence ids, we should print it out.
StoreFile.toStringDetailed is a candidate for that change, there might be more
{quote}
StoreFile.toStringDetailed already display the sequenceId under certain
conditions. I will remove the condition to make sure we always have this
information.
{quote}
What's up with the commented out code in HRegion?
{quote}
Wow, this should not be there. At all! Removed!
{quote}
Is it passing all the unit tests? A trunk version would here to get some Hadoop
QA love.
{quote}
Yes, it's passing unit tests. I will post the results before EOD. Also, there
is no trunk version because it's a backport of HBASE-6630 which is already
doing the same thing in trunk.
So, waiting for your recommendation regarding Durability.USE_DEFAULT tests and
I will post an updated version.
> Cells cannot be overwritten with bulk loaded HFiles
> ---------------------------------------------------
>
> Key: HBASE-8521
> URL: https://issues.apache.org/jira/browse/HBASE-8521
> Project: HBase
> Issue Type: Bug
> Affects Versions: 0.92.1
> Reporter: Jonathan Natkins
> Attachments: HBASE-8521.diff, HBASE-8521-v0-0.94.patch,
> HBASE-8521-v1-0.94.patch, HBASE-8521-v2-0.94.patch, hfileDirs.tar.gz
>
>
> Let's say you have a pre-built HFile that contains a cell:
> ('rowkey1', 'family1', 'qual1', 1234L, 'value1')
> We bulk load this first HFile. Now, let's create a second HFile that contains
> a cell that overwrites the first:
> ('rowkey1', 'family1', 'qual1', 1234L, 'value2')
> That gets bulk loaded into the table, but the value that HBase bubbles up is
> still 'value1'.
> It seems that there's no way to overwrite a cell for a particular timestamp
> without an explicit put operation. This seems to be the case even after minor
> and major compactions happen.
> My guess is that this is pretty closely related to the sequence number work
> being done on the compaction algorithm via HBASE-7842, but I'm not sure if
> one of would fix the other.
--
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
