jacob miller wrote:

Phil,

Am looking for availability reports,bandwidth usage,alerting service and ability to create different logins to users so they can access diff objects

For all in one, OpenNMS does decent and may meet your needs. We often utilize a mixture of tools and modify for working with what we want. My only issue with OpenNMS was that it's java and I don't care to add java to the list of languages I program in. My only complaint was it could get really weird when you have 3,000 unnumbered interfaces. :) Jack

On Thu, Aug 19, 2010 at 7:37 AM, Scott Berkman <scott@sberkman.net> wrote:

I'd recommend ZenOSS.

       -Scott

+1 -B

-----Original Message----- From: jacob miller Sent: Thursday, August 19, 2010 4:36 AM To: nanog@nanog.org Subject: Re: Monitoring Tools

Phil,

Am looking for availability reports,bandwidth usage,alerting service and ability to create different logins to users so they can access diff objects

Thnks,

Jacob, I have not yet found a monitoring environment to my liking and I have seen most of them over the years. That is a project that could keep someone busy for a decade or so (and is one of the things I might work on when I retire). It seems that the more configurable they are, the less intuitive they are and more difficult to get configured properly. Many of the open source tools have only one or two active developers who also have lives outside the project and dealing with a flood of feature requests from the field can be more than they can reasonably accommodate. The commercial monitoring environments can be extremely expensive and very difficult to configure. More important than configuring them is maintaining that configuration over time as things change. I have seen many monitoring environments installed and configured only to become somewhat useless and disused over time as the configuration isn't kept up to date. Good luck in your search but in my experience it generally comes down to putting together a hodge-podge of various tools that give a specific operation the information it needs as those needs vary from one operation to the next. One problem, too, with these tools is that they often collect duplicate information. It would be nice to have some common collector/store so that other tools can pull the information out of that store. Why have three different tools querying snmp stats from the same devices? Having one collector and sharing the data would be a better approach. There is an attempt to consolidate various open source tools in a common framework called GroundWorks. They aren't completely there yet but I believe they are pointed in the right direction. George

Too much widsom in just a single email Paolo On Thu, Aug 19, 2010 at 09:04:13AM -0700, George Bonser wrote:

...

-----Original Message----- From: jacob miller Sent: Thursday, August 19, 2010 4:36 AM To: nanog@nanog.org Subject: Re: Monitoring Tools

Phil,

Am looking for availability reports,bandwidth usage,alerting service and ability to create different logins to users so they can access diff objects

Thnks,

Jacob,

I have not yet found a monitoring environment to my liking and I have seen most of them over the years. That is a project that could keep someone busy for a decade or so (and is one of the things I might work on when I retire). It seems that the more configurable they are, the less intuitive they are and more difficult to get configured properly. Many of the open source tools have only one or two active developers who also have lives outside the project and dealing with a flood of feature requests from the field can be more than they can reasonably accommodate. The commercial monitoring environments can be extremely expensive and very difficult to configure. More important than configuring them is maintaining that configuration over time as things change. I have seen many monitoring environments installed and configured only to become somewhat useless and disused over time as the configuration isn't kept up to date.

Good luck in your search but in my experience it generally comes down to putting together a hodge-podge of various tools that give a specific operation the information it needs as those needs vary from one operation to the next.

One problem, too, with these tools is that they often collect duplicate information. It would be nice to have some common collector/store so that other tools can pull the information out of that store. Why have three different tools querying snmp stats from the same devices? Having one collector and sharing the data would be a better approach. There is an attempt to consolidate various open source tools in a common framework called GroundWorks. They aren't completely there yet but I believe they are pointed in the right direction.

George

On Aug 19, 2010, at 6:23 AM, Phil Regnauld wrote:

jacob miller (mmzinyi) writes:

...

Am looking for an opensource network monitoring tool with ability to create different views for different users.

Hi Jacob,

What kind of network monitoring ? Bandwidth utilization, service availability, RTT, statistics data collection, ... ?

There are tons of open source software tools out there:

Nagios (www.nagios.org) Zabbix (www.zabbix.com) OpenNMS (www.opennms.org) ZenOSS (www.zenoss.com) SmokePing (http://oss.oetiker.ch/smokeping/) Cacti (www.cacti.netl) NetFlow Dashboard (http://trac.netflowdashboard.com/netflowdashboard/) NFSen (http://nfsen.sourceforge.net/)

etc...

Depends on what you want to achieve!

Yes, yes it does... This is a (dated, but still good) introduction that you might want to read: http://www.nanog.org/meetings/nanog26/presentations/stephen.pdf Joe Abley and Stephen Stuart from NANOG26! W

Cheers, Phil

The last time I looked, my main issue with Zabbix was that it required (or greatly preferred) their proprietary agent on every host. This may have changed. -Scott -----Original Message----- From: Nathan Eisenberg [mailto:nathan@atlasnetworks.us] Sent: Thursday, August 19, 2010 2:53 PM To: nanog@nanog.org Subject: RE: Monitoring Tools

Am looking for an opensource network monitoring tool with ability to create different views for different users.

Regards,Jacob

Just to add another opinion to the pot, I've used zabbix in several large environments, and I like it a lot. The developer team is decently sized, and very responsive to requests and feedback (they operate a commercial 'support' model for the platform, so working on the system is literally their day job - as George pointed out, this is often a problem). Zabbix also supports distributed monitoring, which is very handy for scaling or for monitoring multiple locations without dealing with VPNS and the like (or if you have places you need to monitor behind NATs!). Its major weakness at the moment is the weak support for SNMP traps (works great in polling mode, though), so you will want a separate simple system for catching traps. In my opinion, that's just fine, because statistics/trending/basic resource alerting/etc are best kept separate from things like "OMG one of my powersupplies is dead!!11one". Also supports IPMI, which is nice if you have IPMI deployed. :-) Best Regards, Nathan Eisenberg

Nathan Eisenberg (nathan) writes:

It hasn't really changed. Almost every monitoring package I've found where you want to monitor something like 'disk space free on /' requires a daemon of some sort on the host - whether that's SNMPD or their agent.

Anything else than SNMP is a hassle (IMHO). I understand the idea of having a dedicated agent for some hosts - Windows for instance, when querying the WMI - and often it's the only way for a vendor to have a predictable, verified element in the greater scheme (the network). But in most cases, monitoring can be achieved by extending the SNMP mib, and using and custom scripts that will report on mail queue size, in-house application status, etc...

FWIW, I have had their agent running on many, many servers over the years - it has never caused me a moment of heartache (for safety's sake, iptables restricts who can talk to the agent, which has its own control mechanism built in to define who it will talk to, and it runs as a restricted user, just in case).

<hat employer=other> While developing our own monitoring product, we've had to deal with various constraints from the customer side, for instance pharmaceutical companies where there was no way installing an agent on PLC machines would pass internal audit, without having the entire system re-validated (we're talking FDA-validated medication production here). </hat> But often, SNMPD ships with or is available as an optional base component (Windows, most UNIXes) and it's easier to convince the IT suits. Go figure. Oh, and it avoided us having to install an agent on 1000+ servers :) Cheers, Phil

Curtis Maurand (cmaurand) writes:

...

Oh, and it avoided us having to install an agent on 1000+ servers :)

But the configuration learning curve for SNMP is very steep indeed.

Doing network monitoring and not understanding SNMP is like, umm, well I fail to come up with an analogy, but you get my drift. :) It's a bullet you'll have to bite at one point.

Vendor MIBs are the worst part of any new monitoring project. It is made even worse when they change them ever so slightly during an upgrade making your free disk space show as -2tb... On Aug 19, 2010 3:47 PM, "Scott Berkman" <scott@sberkman.net> wrote:

Agreed. And it REALLY isn't that complicated. Go spend some time with CORBA or TL-1 and then re-evaluate the learning curve.

SNMP is really very straight forward as a protocol. If a specific vendor's MIB is difficult to understand or use, that is an entirely different matter.

-Scott

-----Original Message----- From: Phil Regnauld [mailto:regnauld@nsrc.org] Sent: Thursday, August 19, 2010 5:14 PM To: Curtis Maurand Cc: nanog@nanog.org Subject: Re: Monitoring Tools

Curtis Maurand (cmaurand) writes:

...

...

Oh, and it avoided us having to install an agent on 1000+ servers :)

But the configuration learning curve for SNMP is very steep indeed.

Doing network monitoring and not understanding SNMP is like, umm, well I fail to come up with an analogy, but you get my drift.

:)

It's a bullet you'll have to bite at one point.

Looking at ZenOSS to compliment our OpenView NNM system. So far has been pretty simple to get up and running and the support community is pretty responsive to questions. We have cacti in our environment and it works great for pulling bandwidth, CPU, interface errors, mem utilization. the reportit plugin in particular is great for reporting bandwidth utilization for business hours. -- Mike On Aug 19, 2010, at 3:03 PM, Scott Berkman wrote:

The last time I looked, my main issue with Zabbix was that it required (or greatly preferred) their proprietary agent on every host. This may have changed.

-Scott

-----Original Message----- From: Nathan Eisenberg [mailto:nathan@atlasnetworks.us] Sent: Thursday, August 19, 2010 2:53 PM To: nanog@nanog.org Subject: RE: Monitoring Tools

...

Am looking for an opensource network monitoring tool with ability to create different views for different users.

Regards,Jacob

Just to add another opinion to the pot, I've used zabbix in several large environments, and I like it a lot. The developer team is decently sized, and very responsive to requests and feedback (they operate a commercial 'support' model for the platform, so working on the system is literally their day job - as George pointed out, this is often a problem).

Zabbix also supports distributed monitoring, which is very handy for scaling or for monitoring multiple locations without dealing with VPNS and the like (or if you have places you need to monitor behind NATs!). Its major weakness at the moment is the weak support for SNMP traps (works great in polling mode, though), so you will want a separate simple system for catching traps. In my opinion, that's just fine, because statistics/trending/basic resource alerting/etc are best kept separate from things like "OMG one of my powersupplies is dead!!11one".

Also supports IPMI, which is nice if you have IPMI deployed. :-)

Best Regards, Nathan Eisenberg

=+=+=+=+=+=+=+=+=+=+=+=+= Michael Gatti cell.703.347.4412 ekim.ittag@gmail.com =+=+=+=+=+=+=+=+=+=+=+=+=