[
https://issues.apache.org/jira/browse/HDFS-7411?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14288272#comment-14288272
]
Colin Patrick McCabe commented on HDFS-7411:
--------------------------------------------
{code}
public static final String DFS_NAMENODE_DECOMMISSION_BLOCKS_PER_INTERVAL_KEY
= "dfs.namenode.decommission.blocks.per.node";
public static final int
DFS_NAMENODE_DECOMMISSION_BLOCKS_PER_INTERVAL_DEFAULT = 500000;
{code}
Key name doesn't match key string
TestDNFencing: can we avoid moving around the "import static
org.junit.Assert.assertEquals;"? This kind of churn makes backports annoying.
TestPendingInvalidateBlock: do we need this import?
{code}
import org.apache.log4j.Logger;
{code}
{{DecomissionManager#activate}}: can we have some debug logs here spitting out
the configuration parameter values?
{code}
LOG.trace("startDecommission: Node {} is already decommission in "
+ "progress, nothing to do.", node);
{code}
Grammar: "is already decomissioning"
{{getInsufficientlyReplicated}}: this function's name suggests a getter, but it
seems to also schedule replication work. How about calling this
{{handleInsufficientlyReplicated}}?
{{kickMonitor}}: this kicks the monitor, but also waits for the it to complete.
That's kind of different from some of our other "kick" functions. How about
calling this {{runMonitor}} or something like that?
What's the rationale for initializing the {{DecomissionManager}} configuration
in {{activate}} rather than in the constructor? It seems like if we
initialized the conf stuff in the constructor we could make more of it
{{final}}?
> Refactor and improve decommissioning logic into DecommissionManager
> -------------------------------------------------------------------
>
> Key: HDFS-7411
> URL: https://issues.apache.org/jira/browse/HDFS-7411
> Project: Hadoop HDFS
> Issue Type: Improvement
> Affects Versions: 2.5.1
> Reporter: Andrew Wang
> Assignee: Andrew Wang
> Attachments: hdfs-7411.001.patch, hdfs-7411.002.patch,
> hdfs-7411.003.patch, hdfs-7411.004.patch, hdfs-7411.005.patch,
> hdfs-7411.006.patch, hdfs-7411.007.patch
>
>
> Would be nice to split out decommission logic from DatanodeManager to
> DecommissionManager.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
