Re: Network/System Monitors

[Alan Johnson]

On Fri, Aug 21, 2009 at 2:30 PM, Chip Marshall  wrote:

> OpenNMS looks interesting to me, I think I'll have to try it out
> sometime when I get spare cycles...
>

There is significant overhead in the initial OpenNMS setup, not unlike HP
Open view, but not quite as bad because there is a strong community around
it.  Perhaps for some one more seasoned with similar tools, it would be
easier, but I still have to learn a lot of lingo.  Also, the last time I
cracked the box was at least a year ago, and only for about 30 minutes,
including install time, so it might have gotten better.  However, training
and consulting on setup is how they make their money, so maybe it has gotten
worse! ;-p

I still believe it is well worth the effort, partularly if you have Java
skills in house.  Even if you don't, you can write pluggins in any number of
languages, assuming you can't find one that fits your needs in the thousands
that are available.

It is sure to scale beautifully.

My belief are based on some reading and this post cast which is a bit dated
at this point, but still a good overview: http://www.twit.tv/floss15

-- 
Alan Johnson
a...@datdec.com
___
gnhlug-discuss mailing list
gnhlug-discuss@mail.gnhlug.org
http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss/

Re: Network/System Monitors

[H. Kurth Bemis]

On Fri, 2009-08-21 at 14:30 -0400, Chip Marshall wrote:
> On August 21, 2009, Tom Buskey sent me the following:
> > Personally, I'd take a well tuned Nagios for alerts and
> > something else for tracking historical data on graphs. I use
> > swatch with multitail to watch logfiles on multiple systems.

Zabbix can monitor both EventLogs and Syslogs, as well as track and
record log events, although I have ever used Zabbix in such a manner.

> 
> I've generally found Nagios to be very good for service
> monitoring and alerting, but scaling can be painful without some sort of
> addon for config management. The distributed monitoring is available,
> but feels kindof hacky to me.
Zabbix supports service monitoring via several methods, including agent
and agent-less options.

> 
> Cacti is good for graphing SNMP data, but I find it to be a
> major pain to get it to graph anything that it doesn't have
> templates for, and crafting new templates tends to be headache
> inducing for me. That could just be my incomplete understanding
> of SNMP MIBs though.
Zabbix give the ability to poll and SNMP capable device, then work with
a specific element to produce historical graphs and reporting.  Graphing
is already built in, also the ability to zoom and examine historical
data is nice.

> 
> For general purpose graphing, I've been tending toward Munin. The
> plugin API is very simple, so it's easy to write new ones in pretty much
> any language you like. The UI isn't as polished as Cacti though.
Zabbix US is PHP, and standalone.  It can be place on another webserver,
pointed at the SQL database and up and running.  Not sure about 

> 
> OpenNMS looks interesting to me, I think I'll have to try it out
> sometime when I get spare cycles...

I tried OpenNMS, but eventually found that Zabbix was already mostly
there for monitoring and reporting, much easier to configure and used
technology I was already familiar with and comfortability using, like
SNMP-WALK, perl, MySQL and PHP.

http://www.zabbix.com
http://www.zabbix.com/screenshots.php
http://freshmeat.net/projects/zabbix

~k

> 
> ___
> gnhlug-discuss mailing list
> gnhlug-discuss@mail.gnhlug.org
> http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss/

___
gnhlug-discuss mailing list
gnhlug-discuss@mail.gnhlug.org
http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss/

Re: Network/System Monitors

[Neil Joseph Schelly]

The last couple posts have reflected my experience as well.  I use Nagios for 
monitoring and alerting.  I use Munin for graphing and visual representation 
of things.

I'm currently upgrading from Nagios 2 to Nagios 3 and I admit I was kinda 
hoping that I'd see more features in Nagios that would let me combine this 
functionality, but it seems Nagios 3 has focused far more on scalability and 
performance in larger installations.  Perhaps that will help it resolve some 
of the issues some of you have had with it in the past at least.

For the meantime, I'll stick to Nagios and Munin though.  Oh, and MRTG for 
switches and routers. For some reason, while I'm sure I could use Munin for 
these, I've stuck with MRTG for many years of switch/router throughput 
monitoring.  
-N
___
gnhlug-discuss mailing list
gnhlug-discuss@mail.gnhlug.org
http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss/

Re: Network/System Monitors

[Chip Marshall]

On August 21, 2009, Tom Buskey sent me the following:
> Personally, I'd take a well tuned Nagios for alerts and
> something else for tracking historical data on graphs. I use
> swatch with multitail to watch logfiles on multiple systems.

I've generally found Nagios to be very good for service
monitoring and alerting, but scaling can be painful without some sort of
addon for config management. The distributed monitoring is available,
but feels kindof hacky to me.

Cacti is good for graphing SNMP data, but I find it to be a
major pain to get it to graph anything that it doesn't have
templates for, and crafting new templates tends to be headache
inducing for me. That could just be my incomplete understanding
of SNMP MIBs though.

For general purpose graphing, I've been tending toward Munin. The
plugin API is very simple, so it's easy to write new ones in pretty much
any language you like. The UI isn't as polished as Cacti though.

OpenNMS looks interesting to me, I think I'll have to try it out
sometime when I get spare cycles...

-- 
Chip Marshall 
http://weblog.2bithacker.net/  KB1QYWPGP key ID 43C4819E
v4sw5PUhw4/5ln5pr5FOPck4ma4u6FLOw5Xm5l5Ui2e4t4/5ARWb7HKOen6a2Xs5IMr2g6CM


pgpNWzuOXQzV2.pgp
Description: PGP signature
___
gnhlug-discuss mailing list
gnhlug-discuss@mail.gnhlug.org
http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss/

Re: Network/System Monitors

[Cole Tuininga]

Tom Buskey wrote:
> Exposing the historical graphs to users is a good thing: they can see
> the effect of doubling the number of developers on a system.

For historical system information and stats, I personally really like
munin.  Nice and easy to set up, generates pretty graphs, extremely
extensible...

http://munin.projects.linpro.no/

-- 
Cole Tuininga
Lead Developer
co...@code-energy.com
___
gnhlug-discuss mailing list
gnhlug-discuss@mail.gnhlug.org
http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss/

Re: Network/System Monitors

[Tom Buskey]

On Thu, Aug 20, 2009 at 2:49 PM, Kenny Lussier  wrote:

> Hi All,
>
> We are currently using Nagios for monitoring systems and some network gear.
> However, we have found that it is a little lacking in a few areas:
>
> Predictive threshold
> SFlow tracking
> Monitoring several sites from a single data collector.
>
>
> Does anyone have any experience with network/system monitors that combine
> the functionality of Nagios, Cacti, SFlow, etc.? We are looking into
> products from Solar Winds and ManageEngines right now, but I would like to
> hear from others what they think? I am open to OSS and commercial products.
>

I've looked at a bunch (Zabbix, Cacti, Nagios, BigSister, BigBrother, MRTG,
Cricket, Orca (Solaris only))and they all have different strengths.

Nagios is great for alerts when a threshold it reached, but logging isn't so
good.

We're using Zabbix where I am  now, but another admin sets the thresholds
and won't change them.  I've gotten good with email filters :-(  I get
alerts that NIS maps got changed twice a day but the timestamps haven't
changed for weeks.  If I could set my own,I'd probably like it alot.

Personally, I'd take a well tuned Nagios for alerts and something else for
tracking historical data on graphs.
I use swatch with multitail to watch logfiles on multiple systems.

Exposing the historical graphs to users is a good thing: they can see the
effect of doubling the number of developers on a system.
___
gnhlug-discuss mailing list
gnhlug-discuss@mail.gnhlug.org
http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss/

Re: Network/System Monitors

[Cole Tuininga]

Are there no Zenoss fans around here?  I use Nagios myself, but thought
Zenoss looked pretty nice and was thinking about giving it a try...

-- 
Cole Tuininga
Lead Developer
co...@code-energy.com
___
gnhlug-discuss mailing list
gnhlug-discuss@mail.gnhlug.org
http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss/

Re: Network/System Monitors

[Brian Chabot]

Alan Johnson wrote:
> On Thu, Aug 20, 2009 at 4:23 PM, Lloyd Kvam  > wrote:
> 
> http://www.intermapper.com
> is a local product that is worth considering.
> 
> 
> I looked at Internapper years ago when I was running a Wireless ISP and
> it looks very nice at the time.  Still, I also suggest you take a hard
> look at OpenNMS.  It has the power or HP Openview, without the price,
> but with some, but not all, of the setup headaches.  Compare Intermapper
> pricing to have the OpenNMS commercial support help you set it up.

I worked with Alan part of that time including the roll-out of
InterMapper.  I'll second his positive review.  I found it incredibly
easy to install and run.

I have worked with Nagios before and while a good system, I found it a
little clunky at the time, though it may have improved since then.

While I don't have much first-hand experience with OpenNMS (I only
played around with their online demo a couple years ago), I've heard
many network ops people rave about how well designed it is.

I'd say definitely look into both OpenNMS and InterMapper and evaluate
which one fits your needs best.


Brian

-- 
---
| br...@datasquire.net Proprietor: http://www.JustWorksNH.com |
| Computers and Web Sites that JUST WORK  |
|   Work: +1 (603) 484-1461Home: +1 (603) 484-1469|
---
___
gnhlug-discuss mailing list
gnhlug-discuss@mail.gnhlug.org
http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss/