On Thu, 2007-01-04 at 11:57 -0200, Christopher G. Stach II wrote:
> Tim Post wrote:
> > Henning is 100% correct. But there are easier ways to do this, not using
> > snmp and or ssh from dom-0 -> guest or vice versa, See below.
>
> Does it integrate seamlessly with fully featured monitoring and
> ticketing systems, almost all of which use SNMP and SNMP agents that
> come integrated with standard system packages?
Absolutely not, its meant simply to produce xml that can be parsed to
fit pretty much any need. Its for people who want to build their own
solution.
> :P Shit. Even Nagios
> would be better.
No, actually it wouldn't. Nagios is an over-malloc()'ing hog that uses
archaic system() calls to vmstat and other utilities (which also malloc
way more than they need), the thing I linked to is 7k stripped and rides
in dentry (all file handles).
One should not raise loads and eat up contiguous cache to do back end
monitoring :)
> Maybe you should turn it into a Nagios plugin. :)
Why would I want to do that? Its defeating the purpose of using low
level C and simple file handles. Again, its stupid to raise loads and
eat up contiguous cache to monitor a server, such is the case with snmp
and nagios.
I suppose its just a matter of personal preference, and fundamental
understanding of how Linux uses memory :)
Best,
--Tim
_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-users
|
Home