hbase-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Ted Yu <yuzhih...@gmail.com>
Subject Re: Region servers down...
Date Tue, 24 Aug 2010 16:09:41 GMT
It would be beneficial to separate the RS on 192.168.158.179 onto another
machine.

2010/8/23 xiujin yang <xiujinyang@hotmail.com>

>
> Hi
>
> My cluster is in this way.
> Hadoop & HBase are deployed on different machine.
> HBase use the hdfs of Hadoop.
>
> Machine
>
> 4 CPU & 6 G RAM
>
>
>
> Hadoop
>
> 192.168.158.171
>
> 192.168.158.172
>
> 192.168.158.174Send
>
> 192.168.158.177
>
> 192.168.158.176
>
> 192.168.158.180
>
>
> 192.168.158.186
>
>
>
>
>
>
>
> HBase
>
> 192.168.158.179  HMaster & RS
>
>
>
>
> 192.168.158.187  RS
>
>
>
>
>
> 192.168.158.188  RS
>
>
>
>
>
>
>
> At first,we deployed all machine Hadoop & HBase & Mapreduce, we found it
> was easy to make RS down because of memory problem.
> Task of Map/Reduce will eat too much memeory. And Hbase need to use swap.
> So we divided them.
>
> Is this the reason? Or memory is tow small?
>
>
>
>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message