hbase-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Himanshu Vashishtha <hvash...@cs.ualberta.ca>
Subject Re: Replication Configuration Multiple Peers, Which Correct?
Date Fri, 15 Mar 2013 17:39:38 GMT
>  1                     zk1,zk2,zk3,zk4,zk5:2181:/hbase
This is correct.

You only want one peer, right? So, adding separate zk servers as
different peers is not the right way. Read up some of the replication
blogs we have; that should help.

Thanks,
Himanshu

On Thu, Mar 14, 2013 at 8:13 PM, Time Less <timelessness@gmail.com> wrote:
> If I have HBase replication set up, I can do this:
>
>> list_peers
> PEER ID                CLUSTER KEY
>  1                     zk1,zk2,zk3,zk4,zk5:2181:/hbase
>
> Or I can do this:
>
>> list_peers
> PEER ID                CLUSTER KEY
>  1                     zk1:2181:/hbase
>  2                     zk2:2181:/hbase
>  3                     zk3:2181:/hbase
>  4                     zk4:2181:/hbase
>  5                     zk5:2181:/hbase
>
> Both appear to result in functional replication, but we have our suspicions
> that one or the other is incorrect in some subtle way we have not yet
> considered.
>
> Questions:
>
>    1. Is it valid to add the entire Zookeeper ensemble as the peers for
>    replication? Or should we only be specifying one node?
>    2. If the latter, does that mean replication dies if the single ZK node
>    dies?
>    3. If the former, which is the correct way to define the ensemble? The
>    single-peer method, or multiple-peer method?
>
> --
> *Tim Ellis: *Fifth Sigma, Inc. Multimedia and Technology++

Mime
View raw message