[mrtg] Re: Maximum number of nodes that can be monitored with MRT G

[Jerry Heidtke]

David,

It's better to talk about targets rather than nodes or devices. If you're
monitoring a switch, are you going to only monitor traffic on the uplink
port (1 target), or traffic on each port, as well as interface errors,
CPU/backplane utilization, temperature, etc. (potentially 100+ targets for
one node).

That said, I'm monitoring 200 nodes, with 6,000 active targets, updated
every five minutes. Mrtg is running on a Compaq server with a single 1.4Ghz
P3, 1GB ram, Windows 2000. Average CPU utilization as measured by mrtg is
12%; less than half of this is from mrtg (the server is also running Compaq
Insight Manager 7, HP Web JetAdmin, Big Brother, Compaq Version Control
Repository Manager, Compaq Active Update, and a few other services related
to systems management, including two different sql databases).

I figure I could scale this to 20,000 targets without breaking a sweat (and
might reach this later this year). Beyond that, I'd add another CPU and some
more memory. If I needed to go beyond 50,000 targets, I'd probably look at a
different/dedicated platform, or multiple monitoring servers with a single
web server front end.

In short, I think mrtg will handle your expected workload just fine. You may
need to use rrdtool on the backend, and Apache/mod_perl/14all (or similar)
on the front end, but these are easy.

Jerry

-----Original Message-----
From: Bush, David A (IT) [mailto:BushDA at corning.com]
Sent: Thursday, November 21, 2002 1:38 PM
To: 'mrtg at list.ee.ethz.ch'
Subject: [mrtg] Maximum number of nodes that can be monitored with MRTG



Hey folks, 
	Can anyone tell me what the maximum number of nodes is for MRTG?
We're using Trend for reporting right now, and frankly since HP took it over
and we upgraded to the latest version I've --HATED-- it. Everything that was
good about Trend was removed, and on top of everything else it's now about
10 times slower than the previous version. As such - I'm trying to get MRTG
in the door, and my boss is open to the idea. 

	We're currently monitoring about 650 devices and I could see that
roughly doubling in the future. Could MRTG realistically be used to monitor
1300 devices? What kind of hardware would we want for this? I'm assuming a
*nix OS is preferred, which could also help me sneak Linux in the door. 

	The reason I ask is that one of the guys in network management heard
somewhere that MRTG will only scale to about 200 devices. I certainly hope
that isn't the case because I like MRTG and would love it if we could switch
to that. 

Thanks everyone, 
- Dave

Dave Bush
Network Analyst
Corning, Inc.
(607)974-0103
bushda at corning.com


--
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


Confidentiality Notice: This e-mail message, including any attachments, is 
for the sole use of the intended recipient(s) and may contain confidential 
and privileged information.  Any unauthorized review, use, disclosure or 
distribution is prohibited.  If you are not the intended recipient, please 
contact the sender by reply e-mail and destroy all copies of the original 
message.


--
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