[
https://issues.apache.org/jira/browse/HBASE-5746?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13629180#comment-13629180
]
Sergey Shelukhin commented on HBASE-5746:
-----------------------------------------
This is called on context in encodeDataBlock, called from diskToCacheFormat.
Context is then passed around (fixed a little bit in v1, where the same context
was used to set and pass by accident (set called on field but the method arg
passed on, these are currently always the same) to actually use the same one)
until it reaches an implementation of DataBlockEncoder, which calls
prepareEncoding, and gets output stream from it. At that time (because
prepareEncoding was called), dummy header has already been written to that
stream. Previously it would write a header passed at construction time, now it
can be changed based on the version of the block.
> HFileDataBlockEncoderImpl uses wrong header size when reading HFiles with no
> checksums (0.96)
> ---------------------------------------------------------------------------------------------
>
> Key: HBASE-5746
> URL: https://issues.apache.org/jira/browse/HBASE-5746
> Project: HBase
> Issue Type: Sub-task
> Components: io, regionserver
> Reporter: Lars Hofhansl
> Assignee: Sergey Shelukhin
> Priority: Critical
> Fix For: 0.95.1
>
> Attachments: 5720-trunk-v2.txt, HBASE-5746-v0.patch,
> HBASE-5746-v1.patch
>
>
--
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
