You are viewing a plain text version of this content. The canonical link for it is here.
Posted to notifications@accumulo.apache.org by "Josh Elser (JIRA)" <ji...@apache.org> on 2013/12/02 18:24:36 UTC
[jira] [Updated] (ACCUMULO-1947) dfs.datanode.synconclose check is
lacking
[ https://issues.apache.org/jira/browse/ACCUMULO-1947?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Josh Elser updated ACCUMULO-1947:
---------------------------------
Priority: Blocker (was: Major)
> dfs.datanode.synconclose check is lacking
> -----------------------------------------
>
> Key: ACCUMULO-1947
> URL: https://issues.apache.org/jira/browse/ACCUMULO-1947
> Project: Accumulo
> Issue Type: Bug
> Reporter: Josh Elser
> Assignee: Josh Elser
> Priority: Blocker
> Fix For: 1.5.1, 1.6.0
>
>
> The dfs configuration checks we now perform to support hdfs WALs includes a check for dfs.datanode.synconclose which informs the datanode to flush to disk when a file is closed instead of lazily writing the block (very important in the case of power failure).
> Looking at the actual check, it performs reflection on a CreateFlag class. In Apache Hadoop 1.2.1, this class doesn't exist yet the option for dfs.datanode.synconclose does exist. We should likely just reflect on the constant DFSConfigKeys.DFS_DATANODE_SYNCONCLOSE_KEY or DFSConfigKeys.DFS_DATANODE_SYNCONCLOSE_DEFAULT to determine if that option is present.
--
This message was sent by Atlassian JIRA
(v6.1#6144)