Right and AN is not using ZK for the actual NameNode
methods/functions, only for the failover election of the 'backup'
On Sun, Apr 24, 2011 at 4:06 AM, Ozcan ILIKHAN <[EMAIL PROTECTED]> wrote:
> I have read Dhruba's blog entry about AvatarNode long ago. If I am not
> wrong initially it was not using ZooKeeper. That is way I raised that
> question. Also, AvatorNode is build on HDFS 0.20 but my intension is to
> utilize BackupNode that comes with HDFS 0.21
> -----Original Message----- From: Jason Rutherglen
> Sent: Saturday, April 23, 2011 1:17 PM
> To: [EMAIL PROTECTED]
> Subject: Re: HDFS + ZooKeeper
> This is an interesting question. AvatarNode doesn't use Zookeeper?
> Is there a drawback to implementing ZK for the NameNode other than the
> potential bugs from a newly engineered feature?
> Perhaps this issue is the main one for implementing the NameNode in
> ZK? HDFS-1623
> On Fri, Apr 22, 2011 at 10:41 PM, Jean-Daniel Cryans
> <[EMAIL PROTECTED]> wrote:
>> This is a 1M$ question. You could start thinking about this problem by
>> looking at what "AvatarNode" does:
>> On Fri, Apr 22, 2011 at 10:17 PM, Ozcan ILIKHAN <[EMAIL PROTECTED]>
>>> Does anyone have any idea about how we can use HDFS with ZooKeeper? More
>>> elaborately if NameNode fails DataNodes should be able to retrieve
>>> of new NameNode from ZooKeeper.
>>> Ozcan ILIKHAN
>>> PhD Student, Graduate Research Assistant
>>> Department of Computer Sciences
>>> University of Wisconsin-Madison