I have looked at this thread with some interest. Comments below:
M. selcuk karaca wrote:
> I am not a ITIL expert.. In fact a baby on that field.. In this
field I have studied bussiness process Add ON. and I can share my
insights with you..
First (cynical) comment: IME (over many years :) managers (I am now a
manager as well so no insult intended :) will always happily ask for
more info, more detail. When they get it, 90% of the time they do not
use it after the first three weeks.
I think the requirements you are asking for is:
* an overall view of a business service, with a health indication
based on some complex logic to determine the "actual" state
Put the IT componenets, which business service depends on, side by side .
If when that component fails, the bussiness service fails then put an AND
operation after that component. When occurs, Mark business service as CRITICAL
If when that component fails, the bussiness service DOES NOT fail then put an
OR operation after that component. When occurs, Mark business service as
DEGRADED
<snip>
So this would satisfy your customers...
YES!
Assuming this is correct, how would you configure this? I can see
you'd need to find...
Having done the cynical bit, I am not opposed to the suggestion. I have
looked myself into implementing such information systems.
Nagios and thus Opsview is, in my understanding, at heart a monitoring
framework and nothing else. It has no semantic values attached; it does
not know the meaning of the data it handles, it's up to the user to
supply/apply this.
This is its great strength (and at (rare) times, significant weakness).
What this proposal requests is the layering of a semantic level onto the
data (IMU). I've looked at the Business Process Addon (and this:
http://www.netways.de/uploads/media/Bernd_Stroessenreuther_Nagios_Business_Process_View_Addon.pdf
)
and that would be the issue I have with it.
Is this what the monitoring framework should do?
I'd suggest an alternate approach: to enable the exposure of sufficient
and timely data through the API to enable this logic to be implemented
externally (assuming that that is not already possible).
Also I'd suggest a mechanism for permitting the secure gluing in of
results from that external logic (in the form of web pages) into the
existing web interface, as well as the capability to expose separately
individual pages (extending the current capability).
The logic behind that suggestion is to enable a single interface to
present data from multiple external views related to the monitoring
data. Why? Well, the difficulty we find with O-S components is that it's
easy to end up with a massive web portal to cover all the different
service needs of a company/corporate (we are small, but already have
separate Support Desk, NOC 1 (Nagios), NOC2 (Opsview), NOC3 (Nagios),
Webmail, Exchange, Collaboration, CRM, etc etc etc).
If you were interested in doing this, it would permit a general
extensibility without you lot having to create every bit of it, and the
code can stay nicely isolated as well.
OTOH, I think the logic requirements laid out by the OP could however
largely be met by Nagvis (at least to the Pareto level!).
--
James Roberts
Stabilys.com
_______________________________________________
Opsview-users mailing list
[email protected]
http://lists.opsview.org/lists/listinfo/opsview-users
