Drill Down Facility in APAN

[Stanley Hopcroft]

Dear Sir,

I am writing to thank you for your letter and conclude my remarks about
this matter with

On Fri, Apr 25, 2003 at 10:57:11AM -0500, Carroll, Jim P [Contractor] wrote:
> 
> Interesting.  I've made no changes to Outlook.  I don't know why it's
> not wrapping.  (I manually wrapped this, just in case.)
>

(readable bliss!)
 
> I'm not sure what you mean by this.  I'm trying to envision a useful
> host which isn't networked, and am having some difficulty.
> 

Sorry, what I meant was the front-ends whose main function is not
monitoring routers and switches but any old time series such as cache
hit rate, page miss rates, or mails per day eg Orca, but there are
others.

Orcallator could well be useful to you if you only want graphs. It will
collect and graph free mem, numbers of processes, swap rates, etc etc.

Obviously this is useful for capacity planning on your Solaris/Linux
servers (Orcaware is POSIX only I think).

Again, the challenge is that Mr Baddeley has been mentioning: how do you
get Nag to report exceptions on this stuff ?

eg swap rate    > max swap rate threshold
   bytes/sec at critical node
                > threshold

An RRD plugin to compare the last entry added to an RRD with a
threshold, or if you are using Cricket (with anomaly detection), check
the FAILURE archive, is straightforward.

In the Cricket case, you don't need to worry about transient
fluctuations because the Holt-Winters stuff is accounting for it.

However, in the former case, where the value is compared with a static
threshold, it seems to me that Mr Baddelys case of having to change the
threshold because of temporary surges or sags, applies. 

In this case, the RRD does not add anything - it seems to me - to Nag,
unless your RRD plugin analyses the data more extensively (by fetching
the last 10 measurements and computing the average perhaps).

Also, what about hundreds of RRDs ?

> > AFAIK, others whose main purpose is graphing anything such as 
> > Orca (and
> > perhaps MRTG) can accept data from other inputs - usually the file
> > system with data in a specific format. 
>

On the matter of MRTG,

From: Tobias Oetiker <oetiker at ee.ethz.ch>

when mrtg is run in rrd mode it will convert any matching log files
to rrd format automagically

tobi


-- 
 ______    __   _
/_  __/_  / /  (_) Oetiker @ ISG.EE, ETZ J97, ETH, CH-8092 Zurich
 / // _ \/ _ \/ /  System Manager, Time Lord, Coder, Designer, Coach
/_/ \.__/_.__/_/   http://people.ee.ethz.ch/~oetiker   +41(0)1-632-5286


> If you mean running a program/script (such as check_nrpe or check_by_ssh)
> to retrieve data remotely, I can see that.  The drawback is that we
> would have Nagios collecting the data, and then RRD solution X collecting
> the data again.  More network traffic, more CPU churn.  If that's the
> only option, then that's the only option, albeit somewhat sub-
> optimal.
> 

You may well choose to do that because 

. you have an intensive RRD collector saving data for graphing (people
are interested in graphs and archiving 'performance' data: you may have
a requirement to archive data or simple prudence says you should. The
archived data will show trends and be useful in ways not forseen of at
the time)

. you have a less intensive Nag either checking the RRDs, accepting
passive service check results from them (with a link to the graphs) and
doing what RRD usually doesn't, namely checking services rather than
collecting data.

> > What Nagios data do you have in mind that an RRD front-end could
> > leverage ?
> 
> serviceperf.log
> 

IIRC, there are not many plugins (4 at most) that save performance data.

Personally, if I want data about performance, I would probably cosider
collecting such data directly with other tools.

As Mr Baddeley wrote, flood ping or smoke ping may be much better ways
of collecting Network perf data than the Nag checks. Why ? Because the
Nag plugins are less efficient. The Nag plugins are designed to check
services of any nature; perf tools can be optimised because they are
basically stand alone.

> 
> $ snmpget -V
> UCD-snmp version: 4.2.5
>

( I think the latest is net-snmp v5.xx. If you really want to suffer,
the encyption and authentication stuff in SNMP v3 addresses all of the
problems of v1, but it requires some understanding and setting up. In
any case, if you want to monitor routers and switches you can't use it
[because no one implemnts it] other than by proxy. If you used the
net-snmp agent however, you could securely manage hosts that the agent
runs on [Linux/Sol/FreeBSD/])
 
>

Those that have some good ideas about this matter that lead to
code will have the last word about RRDs and Nag.

This is not meant to be an RRD will save us thread: The Nag developers
may simply feel that RRDs can provide some advantages in some cases.

Yours sincerely.

-- 
------------------------------------------------------------------------
Stanley Hopcroft
------------------------------------------------------------------------

'...No man is an island, entire of itself; every man is a piece of the
continent, a part of the main. If a clod be washed away by the sea,
Europe is the less, as well as if a promontory were, as well as if a
manor of thy friend's or of thine own were. Any man's death diminishes
me, because I am involved in mankind; and therefore never send to know
for whom the bell tolls; it tolls for thee...'

from Meditation 17, J Donne.


-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
_______________________________________________
Nagios-users mailing list
Nagios-users at lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/nagios-users
::: Please include Nagios version, plugin version (-v) and OS when reporting any issue. 
::: Messages without supporting info will risk being sent to /dev/null