[wplug] mildly OT: RHN & Solaris

[Chester R. Hosey]

On Fri, 2005-08-19 at 10:16 -0400, Poyner, Brandon wrote:

> I really like Nagios, but I think RedHat is trying to take RHN further.
> For instance most Nagios setups only alert you to the problem, you have
> to fix it yourself.  From the looks of things not only can RHN alert you
> that your load average is high on a server, but you can tell RHN to
> repurpose machines to spread the load over multiple machines.  That
> makes huge assumptions that you have a load balancer and appropriate
> storage.  I don't know how automated you can make it, but if you can't
> now perhaps you will be able to soon.

I'd like to see that happen. I haven't yet seen any functionality from
RHN similar to Novell's ability to migrate, say, web serving between
machines without downtime. It seems to me that such functionality would
be required for automated repurposing.

Of course, the service checks would still need to be done. This isn't
changed if and when you take specific action based upon various
conditions. And there's still the reliability factor of RHN.

I'm really looking forward to Nagios 3.0, since it's been said that the
interface will be abstracted more from the core functionality. I'd
imagine that separation will bring flexibility along with it.

Nagios is really neat as it stands, although I'd like to see more
powerful host grouping functionality, including nested groupings. I'd
also like to see the ability to add services to a template which would
cascade to descendants, or to add services to a hostgroup which would
apply to members. And collapsible views. Those would be nice too. Yeah.

I hope I don't give the impression that I'm never happy. I'm always
happy, but I try to be honest about shortcomings since nothing's perfect
but bacon. I *love* the bacon.

Chet