[jira] [Commented] (HDFS-8145) Fix the editlog corruption exposed by failed TestAddStripedBlocks

["Tsz Wo Nicholas Sze (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/HDFS-8145?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14500742#comment-14500742 ] 

Tsz Wo Nicholas Sze commented on HDFS-8145:
-------------------------------------------

- getCommittedBlock should use blockGroupDataBlocks instead of NUM_DATA_BLOCKS.  BTW, let's rename blockGroupDataBlocks to numDataBlocks?

- sblock.getNumBytes() / HdfsConstants.BLOCK_STRIPED_CELL_SIZE should be
{code}
(sblock.getNumBytes()-1) / HdfsConstants.BLOCK_STRIPED_CELL_SIZE + 1
{code}

Let's split this into two JIRAs since two issues are quite different?  We may reopen HDFS-8166 for the DFSStripedOutputStream issue.

> Fix the editlog corruption exposed by failed TestAddStripedBlocks
> -----------------------------------------------------------------
>
>                 Key: HDFS-8145
>                 URL: https://issues.apache.org/jira/browse/HDFS-8145
>             Project: Hadoop HDFS
>          Issue Type: Sub-task
>            Reporter: Jing Zhao
>            Assignee: Jing Zhao
>         Attachments: HDFS-8145.000.patch
>
>
> {{TestAddStripedBlocks}} failed with some editlog corruption. Did some debugging, I can see at least two issues:
> # DFSStripedOutputStream tries to send out an empty packet to close the block even if writing 0 bytes
> # Because of the above, NN tries to close the file. This exposes another bug in {{BlockInfoStriped}}, which writes its data/parity block numbers into the close editlog but do not read them while loading.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)