hadoop-mapreduce-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Azuryy Yu <azury...@gmail.com>
Subject Re: hadoop 2.0 client configuration
Date Tue, 11 Jun 2013 00:27:02 GMT
if you want work with HA, yes , all these configuration needed.

--Send from my Sony mobile.
On Jun 11, 2013 8:05 AM, "Praveen M" <lefthandmagic@gmail.com> wrote:

> Hello,
>
> I'm a hadoop n00b, and I had recently upgraded from hadoop 0.20.2 to
> hadoop 2 (chd-4.2.1)
>
> For a client configuration to connect to the hadoop cluster, in the
> earlier 0.20.2 case,
>
> I had to specify only the,
> fs.default.name and mapred.job.tracker config parameters.
>
> But now, with the HA configuration of hadoop 2. I have a lot of
> parameters in my client configuration.
>
> fs.defaultFS
> dfs.nameservices
> dfs.ha.namenodes.<nameservice-id>
> dfs.namenode.rpc-address.<nameservice-id>.<namenode-1>
> dfs.namenode.rpc-address.<nameservice-id>.<namenode-2>
> dfs.namenode.shared.edits.dir
> ha.zookeeper.quorum
>
> My Question is, do I really need all these configurations in the client?
>
> I'm looking for minimalist client configuration that would let me work
> with the hadoop cluster in HA mode.
>
> Thanks in advance.
>
> Thanks you,
> Praveen
>

Mime
View raw message