• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

need some network monitoring tool suggestions

S

Soybomb

Diamond Member
I work for an ISP and we're trying to come up with some good monitoring tools to use. Any suggestions are more than welcome.

1.) We run windows nt/2000 server with radius software doing verification. I apologize for not having more information on the network, but out NOC won't give us any more. We need a good way to monitor the useage on our various circuits. IE: See that this T1 is passing 80% of its traffic or all of it.

2.) Some sort of network tool that shows some failures. IE: A DNS server quits resolving some domain names, serving some webpages, etc.

I'm familliar with netsaint and such tools, but I wanted to see what you all thought would be best.
 
as far as bandwidth utilization for different circuits i would recommend polling the network devices via SNMP with MRTG to generate graphs of utilization and such. As far services being up or down; there are quite a few good tools. The best are netsaint [www.netsaint.org] and demarc [www.demarc.org] - netsaint has some cool plugins and even supports WAP. demarc also integrates with snort for intrusion detection in addition to doing file integrity / service monitoring on different servers. HTH.
 
I'll second MRTG - It's an excellent tool and free to boot. Produces excellent reports of traffic statistics, a lot of organizations use it for internal and customer reports.

For monitoring up/down services, I'm a big fan of WhatsUp Gold. Pretty cool product. If you're looking to go cheap, Big Sister is a decent freeware system monitor.

- G
 
I'll second Garion for both MRTG and WhatsUp Gold, where i live a saw some ISP and wireless phone providers using WhatsUp Gold
 
Actually we have mrtg right now (well actually NOC has mrtg, they won't allow us tech support people to login because they say that info isn't useful to us). We've been looking at both whats up and Orion by solarwinds. Orion is looking very very sweet!
 


<< well actually NOC has mrtg, they won't allow us tech support people to login because they say that info isn't useful to us). >>



they don`t want YOU to know what THEY are not doing !!
 
Spidey, I agree that eHealth is the way to go - Nothing else can even compare. Buuttt... In this case, he's likely looking for something on the cheap - We paid $50K+ for our eHealth licence - Most ISP's aren't up to shelling out that kind of dough.. MRTG gives 80% of the functionality and is free! Not to mention that Concord runs best on a really big Sun box and MRTG is quite happy on a Linux or NT box.

- G
 
"The simple fact is, a monitor will not provide consistent and useful information regarding the issues that face the Support Dept. everyday. All it will do is tell you when a Circuit, Server or its Service is down hard"

Damn they really don't want us to be able to watch the network. I sure think it would be handy to look at network utilization among other things. Can anyone see any good reason to not allow access to a network monitoring tool? 🙂
 
Not that its that big of a deal but to provide such a real time tool to customers would mean to allow limited access to the routers and devices not owned by the customer. This would include information about public and private strings for SNMP. Now granted there isn't that much you could do with this type of information but it is data that is configured on their routers that their customers do not own.

It is becoming more political the higher your questions get. Even if the risk is minimal their reason could be purley that it is not a support they want to provide. Free? Some ISPs include it as a competitive edge and some include it at a cost. Also peering does take up bandwidth as minimal as it is there is bandwidth being used to do the polling.

There is also that little bit about you knowing more about thier mistakes and catching them on it that I am sure makes them feel its not an investment to them but a perhaps something that will cost them more money to allow their customers to have.

I agree that it should be used but the world is viewed differnt through business and political eyes.
 


<< perhaps something that will cost them more money to allow their customers to have. >>


Hehe you misunderstand, I work for the ISP as tech support. We don't have access to any network monitoring tool and are told we don't need to. I see no reason why tech support cant have this access!
 
If your Tech Support you have absolutely no business monitoring the network. That is what NOC is for.

Sounds like NOC knows best.
 


<< Not that its that big of a deal but to provide such a real time tool to customers would mean to allow limited access to the routers and devices not owned by the customer. This would include information about public and private strings for SNMP. Now granted there isn't that much you could do with this type of information but it is data that is configured on their routers that their customers do not own. >>



If the NOC sets up a tool like MRTG, renders the data, throws it up on a web server and gives this tech support guy the URL, how is that exposing SNMP ports to the rest of the network? He can keep those filtered to whatever the machine is that's actually compiling the MRTG stats. As for the 'need to know' regarding network usage, it seems silly to me to be so secretive about network usage graphs; certainly there isn't a dying need for a tech support guy to check out network stat graphs, but it couldn't hurt to have access to it. What vulnerability can this expose to an organization to allow another department to see what the traffic load is on the network? Couldn't they just walk into the data center and look at the activity lights on the gear anyways? 🙂
 
Sticking the data into a static web page as a graph was not what I was speaking of, but real time monitoering. I do agree though that a static page would be something they could do and not be as worried about the information in their routers.
 
even realtime, what if the webserver is making the query rather than the client system? That still provides the ability to filter SNMP totally except for web server <-> network device. good example == www.demarc.org 🙂
 
Suggestions:

Free/Opensource:

MRTG
Netsaint
Rancid

Commercial:

e-health
Riversoft


Finally, I've never known a company where NOC knows best, NOC is normally clueless.
 
we've got a router here at my school that they have all the students behind. it averages 130ms of latency and 4% packetloss its so overloaded.
One of my friends tried to figure out what kind of router it was, and if some of us could chip in and BUY a replacement, so he did "nmap -ss -o 148.100.1.9"

bam, two days later, kicked off the network for 'trying to hack the router'.

NOC's, like most IT professionals, are usualy cluess wastes of meat (which btw applies to helpdesk too... usualy more often to boot).

As usual there's a trade off, and with anything you setup (be it an MRTG page, or monitoring software) its never as easy as it seems, and someone has to support it. So simply put they may not have the time. But if you ask me if a help desk tech can answer "Yes we're sorry about that. X is down and we're already aware and working on it" is a much better answer than "uhh.... i'll forward your question to the NOC guys"

but of course, the only thing more fun than incompetant IT professionals at work, is incompetant IT proffesionals in teritorial pissing matches.

bart
 


<< but of course, the only thing more fun than incompetant IT professionals at work, is incompetant IT proffesionals in teritorial pissing matches. >>


Amen brother. Our noc is located a few states away so if they don't answer their phone we have to email and wait, we can't just go bang on their door. As a recent example, we had an outage a few days ago that prevented a large number (nearly all) of our customers from being able to connect properly. This outage lasted over 12 hours and during that time we couldn't even get our noc to place a message on the phone system acknowledging the outlage so the tech's had to deal with a huge volume of calls for an unusually long amount of time.

In the perfect isp, no NOC should do all the monitoring and tech's shouldnt have to worry about it. If noc doesn't do any monitoring of network utilization (as they admit they don't), then someone has to.
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top