You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@hbase.apache.org by Wayne <wa...@gmail.com> on 2011/06/06 20:55:13 UTC

Node Monitoring

Are there any recommended methods/scripts to monitor nodes via nagios? It
would be best to have a simple nagios call to check hadoop, hbase, &
thrift separately and alarm if one of them is awol (and not have the script
cause damage like I have read with thrift). For example our friendly CMF
issues which will occasionally go beyond the 60s zookeeper timeout will
cause hbase nodes to shutdown. We want a nagios alarm/text message to be
sent to let us know we have to go restart the region server.

I know we can figure out something on our own...but wanted to see if there
are some std methods for monitoring that have been developed and shared that
we can use.

Thanks.

Re: Node Monitoring

Posted by Jean-Daniel Cryans <jd...@apache.org>.
We do simple TCP check, and I know that some people run a more
extensive check that calls a low overhead method like getConfiguration
just to see if the region server can answer requests.

J-D

On Mon, Jun 6, 2011 at 6:55 PM, Wayne <wa...@gmail.com> wrote:
> Are there any recommended methods/scripts to monitor nodes via nagios? It
> would be best to have a simple nagios call to check hadoop, hbase, &
> thrift separately and alarm if one of them is awol (and not have the script
> cause damage like I have read with thrift). For example our friendly CMF
> issues which will occasionally go beyond the 60s zookeeper timeout will
> cause hbase nodes to shutdown. We want a nagios alarm/text message to be
> sent to let us know we have to go restart the region server.
>
> I know we can figure out something on our own...but wanted to see if there
> are some std methods for monitoring that have been developed and shared that
> we can use.
>
> Thanks.
>