-----Original Message-----
From: Suresh Srinivas (JIRA) [mailto:j...@apache.org]
Sent: Thursday, July 21, 2011 12:40 PM
To: hdfs-issues@hadoop.apache.org
Subject: [jira] [Commented] (HDFS-2167) Move dnsToSwitchMapping and hostsReader
from FSNamesystem to DatanodeManager
[
https://issues.apache.org/jira/browse/HDFS-2167?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13068840#comment-13068840
]
Suresh Srinivas commented on HDFS-2167:
---------------------------------------
# Can we keep BlockManager#corruptReplicas private?
# Is there a jira created to not to use FSNamesystem, FSNamesystem#datanodeMap
FSNamesystem#heartbeats for locking in block manager? If not can you please
create one?
# Should we track with a jira removing references to NameNode.stateChangeLog in
blockmanagement package to appropritate logger from blockmanagement package?
# Moved code resolveNetworkLocation() no longer has {{assert hasWriteLock()}}.
What is the locking mechanism for this code?
# FSNamesystem#isDatanodeDead(), FSNamesystem#updateStats(),
FSNamesystem#blockTokenSecretManager, FSNamesystem#datanodeMap,
FSNamesystem#heartbeats should move to blockmanagement package.
# getDatanodeListForReport()
#* In jiras moving the code, if the moved code remains the same, it would be
easier to review. Case in point - getDatanodeListForReport()
#* Why are doing adding these two sizes {{final List<DatanodeDescriptor> nodes
= new ArrayList<DatanodeDescriptor>(namesystem.datanodeMap.size() + (mustList
== null? 0: mustList.size()));}}
#* Changing mustList from Map to ArrayList is going to slow down removes.
# checkDecommissionStateInternal() no longer checks for hasWriteLock()?
# DatanodeManager should not have namesystem reference?
# In some tests where you have moved from FSNamesystem methods calls to
BlockManager method calls, how is the locking that was done in FSNamesystem
call done with this change?
> Move dnsToSwitchMapping and hostsReader from FSNamesystem to DatanodeManager
> ----------------------------------------------------------------------------
>
> Key: HDFS-2167
> URL: https://issues.apache.org/jira/browse/HDFS-2167
> Project: Hadoop HDFS
> Issue Type: Sub-task
> Components: name-node
> Reporter: Tsz Wo (Nicholas), SZE
> Assignee: Tsz Wo (Nicholas), SZE
> Attachments: h2167_20110719.patch, h2167_20110720.patch,
> h2167_20110720b.patch, h2167_20110720c.patch
>
>
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
DISCLAIMER
==========
This e-mail may contain privileged and confidential information which is the
property of Persistent Systems Ltd. It is intended only for the use of the
individual or entity to which it is addressed. If you are not the intended
recipient, you are not authorized to read, retain, copy, print, distribute or
use this message. If you have received this communication in error, please
notify the sender and delete all copies of this message. Persistent Systems
Ltd. does not accept any liability for virus infected mails.
