hadoop-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "bit1129@163.com" <bit1...@163.com>
Subject Re: Re: Multinode setup..
Date Sat, 03 Jan 2015 07:17:38 GMT
1. There is a slaves file in the conf dir. Add the slave's IP  in it.
2. Configure SSH for each node to logon the other without password
3. Format the the already running node, with hdfs namenode -format
4. starup the Master
Then, at this point, if you encounter problem, the problem will be specific, which you can
google for answers.

From: Anil Jagtap
Date: 2015-01-03 15:05
To: user@hadoop.apache.org
Subject: Re: Multinode setup..
I have a single node already set as per the pdf instructions. 
I cloned this node and want to set as a slave. 
How do I do this ?

Rgds, Anil

Sent from my iPhone

On 3 Jan 2015, at 18:00, "bit1129@163.com" <bit1129@163.com> wrote:

Setting Haddop in real distributed cluster isn't that complex, Which step were you stuck at?

From: Anil Jagtap
Date: 2015-01-03 14:32
To: user
Subject: Multinode setup..
Dear All,

Im trying to setup a multi node cluster and I found millions of articles on how to configure
multi node. I tried couple of them and i really don't understand why Am i not able to configure
it.. I have a feeling that these experts in their blogs omit out something thinking that its
not necessary to publish since its a small thing and the person who is configuring is good
enough to understand but the fact is there are people like me who are completely new to Hadoop,
Linux & Java as well.

So does any one has some kind of material who literally tells me each and every step on how
to configure single node, then add another node and finally link it.. 

Sorry to ask this granular detail but trust me i had wasted good amount of time in searching
and trying it out.. Attached one i found and this helped me in setting up single node successfully.

I can use Cloudra but it consumes lot of memory and hence i want to configure my own so i
can fit 2 nodes in the same memory what Cloudier uses. Any VM changes to Cloudera stops working.

Finally, for now do I really need all of the components or just Sqoop, PIG, HBASE & Hive
should be sufficient to practise.. 

If anyone can share their multi node VM's would be really great.

Thanks to all in advance and expecting good feedback at the possible earliest.

Kind Regards,

共有 1 个附件
hadoop.pdf(358K) 极速下载 在线预览 
View raw message