On Sat, Oct 30, 2004 at 02:25:02PM +0200, martin f krafft wrote:
> I would like to monitor all the nodes of a cluster, but I am rather
> pressed for time so that I cannot investigate all the options.
>
> I tried spong, but it's pretty bad especially because it requires
> changes to the client to specify which tests to run. Ideally,
> a network monitoring system should consist of a client (running on
> the master), and servers on all nodes, which can then do as the
> client instructs.
How about munin? I mentiond this two days ago on this list, but it does
(more or less) what you suggest, with the exception that it "does not do
as the client instructs", rather, it "gives the numbers as the client
wants them".
A good example of how munin works is at their website:
<http://www.linpro.no/projects/munin/example/>
--
EARTH
smog | bricks
AIR -- mud -- FIRE
soda water | tequila
WATER
-- with thanks to fortune
Attachment:
signature.asc
Description: Digital signature