hadoop-hdfs-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Remi Catherinot (JIRA)" <j...@apache.org>
Subject [jira] [Updated] (HDFS-8001) RpcProgramNfs3 : wrong parsing of dfs.blocksize
Date Fri, 27 Mar 2015 13:26:53 GMT

     [ https://issues.apache.org/jira/browse/HDFS-8001?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]

Remi Catherinot updated HDFS-8001:
----------------------------------
    Target Version/s: 2.6.1
              Status: Patch Available  (was: Open)

affected line depends on hadoop version, here are the diff for 2.5.2 and 2.6.0 versions

hadoop version 2.5.2 diff
182c182
<     blockSize = config.getLong(DFSConfigKeys.DFS_BLOCK_SIZE_KEY,
---
>     blockSize = config.getLongBytes(DFSConfigKeys.DFS_BLOCK_SIZE_KEY,


hadoop version 2.6.0 diff
187c187
<     blockSize = config.getLong(DFSConfigKeys.DFS_BLOCK_SIZE_KEY,
---
>     blockSize = config.getLongBytes(DFSConfigKeys.DFS_BLOCK_SIZE_KEY,

> RpcProgramNfs3 : wrong parsing of dfs.blocksize
> -----------------------------------------------
>
>                 Key: HDFS-8001
>                 URL: https://issues.apache.org/jira/browse/HDFS-8001
>             Project: Hadoop HDFS
>          Issue Type: Bug
>          Components: nfs
>    Affects Versions: 2.5.2, 2.6.0
>         Environment: any : windows, linux, etc.
>            Reporter: Remi Catherinot
>            Priority: Trivial
>              Labels: easyfix
>   Original Estimate: 2h
>  Remaining Estimate: 2h
>
> org/apache/hadoop/hdfs/nfs/nfs3/RpcProgramNfs3.java use Configuration.getLong to get
the dfs.blocksize value, but it should use getLongBytes so it can handle syntax like 64m rather
than pure numeric values. DataNode code & others all use getLongBytes.
> it's line 187 in source code.
> detected on version 2.5.2, checked version 2.6.0 which still has the bug.



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

Mime
View raw message