[mrtg] Re: monitoring Windows NT/2000 servers

[Marlo Montanaro]

As long as BB is running continuously (which is why you put it on
Unix/Linux) it will report how long it has found the server up (which will
not necessarily match the actual uptime of the server, unless you reboot it
while BB is up).  You can also configure it to turn Red when a particular
server reaches a specified amount of time that it has responded to pings-
basically saying "time to reboot."

The other nice thing about BB is that it can test various other services- to
me, the fact that the services are up and running is what is important- not
Windows itself.  Proof in point- yesterday my WINS service died on our older
NT box (something we don't test with BB, unfortunately) which took out many
services on our network- but Windows was still ping-able.  Yes- I'm working
on a way to get BB to monitor WINS now...  there are plenty of 3rd party
scripts available.

But you get the point.

And of course the two other best parts of BB are 1- it's free, and 2- it'll
page you when it finds a problem with a server- something I don't believe
MRTG can do, at least not as easily.

You've already got MRTG running on Red Hat, simple matter to add BB to the
same box.  I have both running on a P-II with 128 MB of RAM on Solaris 8 for
Intel.  The BB/MRTG box runs for months at a time without a hiccup.  I think
the two together provide an excellent picture of network availability and
bandwidth availability/usage.  They also let you respond proactively instead
of reactively.  Put them on an older computer that maybe you had no use for
anymore (i.e., Windows XP is a slow dog on it, but Linux runs great!) and
you'll look great to your boss... leverage older equipment with no
expenditure of funds and get a system that lets you know about all kinds of
problems on your network.  Good job.

Good luck!
MM



-----Original Message-----
From: Allen Crawford [mailto:AllenC at mailcode.com]
Sent: Friday, January 03, 2003 3:08 PM
To: 'mrtg at list.ee.ethz.ch'
Cc: Marlo Montanaro
Subject: RE: [mrtg] Re: monitoring Windows NT/2000 servers


Yeah, I thought about that graphing thing.  I guess the graphs aren't really
important.  The amount of uptime is important, however.  It is an objective
for our department by management to have accurate uptime/downtime numbers.
A graph would either be ON (up) or OFF (down) I suppose, so over time it
would be mostly up hopefully.  But I was thinking if you installed SNMP on
the server that it would be able to report the actual uptime like the
router's do with MRTG.  We have some software currently that can do this,
but it won't let us do automated HTML graphs that we can post online.

Thanks for the tip on Big Brother.  I was actually in the middle of reading
about their product.  Our ISP recommended it as well since they use it.

-----Original Message-----
From: Marlo Montanaro [mailto:mmontanaro at centennialcorp.com]
Sent: Friday, January 03, 2003 12:57 PM
To: 'mrtg at list.ee.ethz.ch'
Subject: [mrtg] Re: monitoring Windows NT/2000 servers


Sounds more like you want a tool to tell you if the server is up (i.e.,
available).  Do you care what the actual up time is?  If you are graphing
uptime, it would (theoretically) just continually rise- there is no upper
limit (well, at least if it is a Novell or Unix/Linux box :-) ) and only
return to zero upon reboot (a sawtooth waveform!)

Sounds like you want Big Brother (http://bb4.com) which runs fine on Red
Hat...

Big Brother will also allow you to monitor various services that your
Windows box may be providing (i.e., POP3, SMTP, HTTP, etc.) and let you know
if they are up and page you if they are not.

There are also 3rd party scripts you can add to Big Brother that allows it
to work with MRTG, so various statistics can be graphed also.

Just a thought...

MM


-----Original Message-----
From: mrtg-bounce at list.ee.ethz.ch [mailto:mrtg-bounce at list.ee.ethz.ch]On
Behalf Of Allen Crawford
Sent: Friday, January 03, 2003 12:15 PM
To: MRTG Mailing List (E-mail)
Subject: [mrtg] monitoring Windows NT/2000 servers


Ok, I know there is lots of information out there on this subject, but it
all seems to be based (or at least the stuff I'm finding) on using the
Windows version of MRTG.  I currently have MRTG running great on my Red Hat
7.3 box and it is monitoring our Cisco router.

What I'd like to do, is play with the Windows NT/2000 SNMP service and
monitor these servers FROM my Linux MRTG, not from a Windows MRTG.  I don't
necessarily care to use the third-party SNMP4NT and SNMP42K stuff.  I also
don't necessarily care to monitor CPU utilization or most of the other
performance counters.  What I do want to monitor and graph with MRTG is
simply the uptime of a server.  I might want to add the other stuff later,
but for now I just want uptime.  Basically, if I can ping it, it is up.
That's as simple as I want to begin with.

So, are there any web pages out there that describe how to set this up that
you have seen?  Any other tips for me?

Thanks a lot,
Allen


--
Unsubscribe mailto:mrtg-request at list.ee.ethz.ch?subject=unsubscribe
Archive     http://www.ee.ethz.ch/~slist/mrtg
FAQ         http://faq.mrtg.org    Homepage     http://www.mrtg.org
WebAdmin    http://www.ee.ethz.ch/~slist/lsg2.cgi

--
Unsubscribe mailto:mrtg-request at list.ee.ethz.ch?subject=unsubscribe
Archive     http://www.ee.ethz.ch/~slist/mrtg
FAQ         http://faq.mrtg.org    Homepage     http://www.mrtg.org
WebAdmin    http://www.ee.ethz.ch/~slist/lsg2.cgi

--
Unsubscribe mailto:mrtg-request at list.ee.ethz.ch?subject=unsubscribe
Archive     http://www.ee.ethz.ch/~slist/mrtg
FAQ         http://faq.mrtg.org    Homepage     http://www.mrtg.org
WebAdmin    http://www.ee.ethz.ch/~slist/lsg2.cgi