Talk:Comparison of network monitoring systems

additional columns

this page was quite a good start in looking for a proper system, in that it helped ruling those out that would never fit. Unfortunately the table is missing information about whether a system can warn about thresholds, and automatically learn such thresholds over time. — Preceding unsigned comment added by 213.61.9.75 (talk) 09:58, 31 May 2011 (UTC)

Network monitoring systems basically use two protocols to retrieve data from remote nodes: SNMP and WMI. Should WMI column be added next to SNMP column? — Preceding unsigned comment added by Sonic1980 (talk • contribs) 11:26, 4 September 2011 (UTC)

General comments

Also, it'd be nice to have a row for 'inventory tracking' or similar, as several NMS's supposedly do that well. Thanks. Bobbozzo (talk) 00:19, 18 July 2008 (UTC)

Not sure the "Response Time" column is adding anything, they are all listed as fast!? This needs to be removed or quantified? Mtcooper (talk) 14:19, 27 July 2008 (UTC)

Removed. --Goldfndr (talk) 09:03, 18 August 2008 (UTC)

Couldn't stand not having the table setup right so that it could be sorted. Fixed it.MikeD123 (talk) 00:07, 6 August 2008 (UTC)

Does this really need more footnotes? --ssd (talk) 15:35, 27 August 2008 (UTC)

I'm not sure if http://www.spiceworks.com/ belongs to this table but if it does, then it would be helpful for people if we add it. 198.102.112.18 (talk) 19:27, 2 September 2008 (UTC)

I think SpiceWorks should be added. It should be noted that the produce is not free, but is Ad Supported. I cannot review it as I'm not prepared to install it on any of my servers. DudleyThorpe (talk) 02:27, 3 September 2008 (UTC)

Zabbix does support "Logical Grouping" if this means putting hosts into groups like "AIX". It also allows you to define a "Service" built from items in the inventory or other services: eg WebSite "Service" = Ordering "Service" + Customer "Service"; Ordering "Service" = Ordering App + Ordering DB; etc It also has a Map function that allows you to graphically represent you IT infrastructure and link it together with color/style coded links. DudleyThorpe (talk) 05:58, 3 September 2008 (UTC)

agreed, "logical grouping" should be explained - does it refer to host groups ? to monitored item groups ? --Richlv (talk) 11:09, 18 September 2008 (UTC)

yes, this is truly confusing. And how can Nagios be said to support "logical grouping" when GWOS does not? GWOS supports at least as much grouping as nagios does, as it is all nagios under the hood... LarsHolmberg (talk) 21:45, 8 November 2008 (UTC)

How about a column for indications of types of specific monitoring support -- e.g., Cisco or Juniper mibs, http/https service, Windows or Unix specific details. A good model for this would be the (almost obsolete) http://en.wikipedia.org/wiki/Cisco_PIX PIX page -- in other words, the column would give summaries like 'many' or 'Unix, network' with superscripts that point to footnotes with more complete details. The chart is also too large as it is, so this might be a good way to collapse categories. Categories in general seem to have sprung up to hype particular features of one app versus others. Another piece of useful information (as I'm finding in looking this month for monitoring packages) would be details on packages that include other packages (e.g., Zenoss, Zabbix -- Nagios ; opsview -- Nagios and MRTG, NMIS (MRTG and NMIS based themselves built on RRDtool) etc.). Also needs info on the server requirements (OS at a minimum), a number of packages now have installers for only specific systems (Ubuntu, Debian, Windows, Solaris, etc.), source for others, and a few now have pre-packaged VMWare images as well (NAI/ziptie, opsview, etc.) Skandha101 • 16:52, 8 July 2009 (UTC)

while i agree that table is a mess, adding more columns wouldn't help. maybe it should be split in 3 tables, like "general info" (like chart, autodescovery, distributed monitoring etc support), "can monitor" (snmp, ipmi, agents, tcp, icmp etc) and "technical info" (language written in, server os, data storage method etc). btw, what did you mean by "details on packages that include other packages (e.g., Zenoss, Zabbix -- Nagios" ? :) --Richlv (talk) 12:57, 9 July 2009 (UTC)

I actually agree that it needs fewer columns, not more (which is what I was trying to say with the " a good way to collapse categories" and footnotes, bit). However, we do need to (and this is part of what you get at I think) prioritize the columns or present the important ones meaningfully. Maybe splitting it out would be the way to go. I recently re-discovered a master list of monitor stuff at http://www.slac.stanford.edu/xorg/nmtf/nmtf-tools.html . It does categorize by core functions (flow monitoring, analyzers/sniffers, host monitors) etc. -- I'm not sure it entirely succeeds (what about overlap or product development/scope increase?), but it is very useful as a starting point for some kinds of categorization.

As far as "packages that includes other packages," I mean that there should be some way to point out that some tools have either a monolithic multi-app install (one install script for a bunch of tools that also have independent existence, for instance) or piece-by-piece method (and even dependency) on different packages for the "one" tool (glorified super-dashboard?) to be set up. ossim is like that, as is opsview. the zabbix, groundworks, zenoss, etc. group all are built on top of nagios. Does that make sense? These apps that collect and simplify configuration (one conf file for multiple apps, e.g.) or set-up are very nice in some ways, and I think this is a good thing to point out to readers/people who may be evaluating different apps/frontends. Thanks. Skandha101 • 23:25, 9 July 2009 (UTC)

"the zabbix, groundworks, zenoss, etc. group all are built on top of nagios. Does that make sense?" - i'm sorry, but it doesn't if only because zabbix is not built on top of nagios ;) --Richlv (talk) 08:36, 10 July 2009 (UTC)

I don't know if it's indended to be, but this article could be a great help for people trying to decide which monitoring tool to choose, or change to if they are not satisfied with theirs. A very important point then, as deciders focus on cost, would be a column saying e.g. "Method (and ease) of setup/configuration". The ease part is, obviously, almost impossible to accomplish.

I wouldn't want to explain to the managers why we have chosen that great, flexible, open source solution and then spent months figuring out the configuration to meet out requirements. 92.79.180.153 (talk) 10:01, 30 April 2010 (UTC)

I agree with the concept, although it seems tough to implement in a meaningful way. "Ease of setup" has the potential to be very arbitrary and subjective. "Method" might be useful -- can you give some examples of the kind of content you'd put in a field like that? Lahnfeear (talk) 13:55, 30 April 2010 (UTC)

Webapp

I think too many meanings are packed into this column. Currently, "Webapp" covers the following:

Web

Does this applicaiton display data on the web (yes for all listed so far)

Web management

Does this application allow managing of monitored items from the web

Events

the ability to acknowledge and record remedial actions

Someone just added the events column. Since all of these apps have web pages, and events is already represented, I'd suggest we reorder (and perhaps rename?) events, and rename Webapp to Web management. --ssd (talk) 12:28, 8 September 2008 (UTC)

Events

I think the 'Alarms / Triggers' column conflicts with the 'Events' column, maybe rename 'Events' to 'Event Management', i.e. being able to assign events, accept events, acknowledge/close events, filter events, etc. —Preceding unsigned comment added by 24.5.14.40 (talk) 04:22, 12 September 2008 (UTC)

Is events fully redundant? Surely a system doesn't have triggers, but no way to acknowledge them?? Counterexample? --ssd (talk) 04:37, 14 September 2008 (UTC)

This article is an Advertisement

Or is there some other reason for i-enable rmf to be bold red in the product column and have a favorable result in each category? I went to the website for the product (which is only an ip address by the way, no domain name) it belongs to 3i Infotech, the company that makes the product, where they are selling remote managed services (using their rmf product).
If the article is not going to be deleted, I would at least recommend deleting the row for i-enable.Billybennett (talk) 18:15, 10 June 2009 (UTC)

entry is red because the linked page does not exist. quite basic mediawiki functionality. now, if that i-enable data is not correct, that is a different matter - it should be corrected or the row indeed deleted. --Richlv (talk) 06:49, 16 June 2009 (UTC)

Inventory

Someone add a description what "inventory" should include, and I'll fill it in for the apps I know... --ssd (talk) 12:14, 23 October 2008 (UTC)

What exactly does "inventory" mean for this? —Preceding unsigned comment added by 192.207.234.194 (talk) 17:28, 17 April 2009 (UTC)

Extensions (External Scripts, Plugins, Plugin Creation)

The "External Scripts", "Plugins", and "Plugin Creation" columns are ambiguous. I propose that "Plugins" only refers to modules (shared objects, Perl-modules, ...) that are loaded at runtime and stay in memory but not the ability to execute custom scripts. --octo 131.188.30.161 (talk) 16:21, 10 November 2008 (UTC)

The description of the "Plugin Creation" column is not helpful: what should go into that column? The general difficulty to write a plugin? Do custom scripts count, too? What if multiple possibilities exist? I propose to change that column to "no/yes + short description", for example "yes, .NET assemblies", "yes, embedded Perl", "no". The contents of "External Scripts" should not be considered for "Plugin Creation", given that external scripts are easy to write for any reasonably skilled systems administrator. --octo 131.188.30.161 (talk) 16:21, 10 November 2008 (UTC)

Why does this article exist?

Isn't it synthesis (I don't see any links to reliable sources comparing all these products using these criteria) and original research (e.g. how were the comparison criteria chosen)? Please can someone explain how including this article (and presumably any number of other "Comparison" articles) can be justified? --Rogerb67 (talk) 01:21, 25 January 2009 (UTC)

A: I'm not a wiki editor, I'm just a wiki user. So I don't know about synthesis or other criteria you use in deciding if the article is good or not. I just know that I find this document useful at least because it brings to my attention all the different tools out there. It saves money for my company as I can look if there are free tools versus the tools I pay for and so on. Please rethink the deletion status. Again, I find it useful and I'm not the only one. Of course, if you want to make it more reliable please do it. —Preceding unsigned comment added by 82.196.214.14 (talk) 10:19, 26 March 2009 (UTC)

collectd and trending

Per their site, collectd "...does not generate graphs. It can write to RRD-files, but it cannot generate graphs from these files." This software simply collects statistics and stores them in different formats (some traditionally used for trending, such as RRD). It is NOT intrinsically capable of trending. Lahnfeear (talk) 00:50, 5 February 2009 (UTC)

ZENOSS VERSUS SCOM

Hi,

Could you please help me in providing the following comparison report?. We are basically looking for a comparison, advantages and dis-advantages report between Zenoss monitoring tool and SCOM (Zenoss VS SCOM).

Zenoss :

Network monitoring YES

System monitoring YES

Database monitoring YES

Application monitoring YES

Middleware monitoring YES

SLA tracking YES

Multitenant accommodation YES

Software asset monitoring YES

Network configuration monitoring YES

Performance monitoring YES

Storage tracking YES

Forecasting Yes

Storage capacity planning YES

Reporting Yes

Integration with third-party tools YES

NEW Cloud Management Service (CMS) YES —Preceding unsigned comment added by Raj esh90 (talk • contribs) 17:28, 26 May 2009 (UTC)

Free versus non-free, restrictions

I'd like to see more of a distinction between free solutions and commercial solutions, perhaps even breaking out products into (at least) two categories, as the people who are looking are FOSS are frequently not looking at commercial enterprise software, and vice versa. This is complicated when there is something like a 'community'/free edition and an 'enterprise'/paid version. In those cases it would be good to know what the limitations are on the community edition (e.g., monitor up to 100 nodes before licensing is necessary, etc.). Skandha101 • 16:52, 8 July 2009 (UTC)

what's "user tracking" ?

recently a new column, named "user tracking" was added with only few data points, and cisco product being the only with "yes". the already huge table now is even less readable, and there is no explanation offered as to what the hell does that even mean. it would be better to first discuss adding new columns on this talk page, and they should be explicitly explained, including why it is a noteworthy enough feature to be added to the table. once the table has been split, we can add whatever feature next monitoring system decides to advertise, but until then i'd suggest against adding new columns. unless a good rationale is provided for "user tracking", i vote for removing it, at least until tables is split up. --Richlv (talk) 10:31, 27 July 2009 (UTC)

omission of major commercial packages ?

Why no comparisons of these products to large commercial packages like HP Openview, Netcool, Tivoli, etc?

Bmccleer (talk) 08:12, 6 October 2009 (UTC)

omission of major commercial packages and ambiguous deletion of new vendors

BMcclear - there should be no reason why large commercial products can't be added by their respective vendors--- and if there is a feature in question, it should be called out in the talk page. In the same manner --- there should be no reason why other new commercial products, such as AccelOps, can't be added. For some reason, different vendors are being arbitrarily deleted. This is a very useful chart that many refer to. It provides a basic overview... and the means for users to investigate further on their own. I am sure opnet and others would agree.

Scottgwiki (talk) 03:25, 11 December 2009 (UTC) Scottgwikip (talk) 00:01, 12 December 2009 (UTC)

actually, this table is crap. first, columns are vague. second, there are too many of them. third, some of them are not explained. wtf user tracking ?

so here's what i would propose. 1. split the table in functional groups, creating several tables; 2. remove crap, undefined columns - probably added by some marketoid. if nobody knows what something means, just nuke it. 3. create specific definitions of columns; 4. no new columns added w/o discussion; 5. all cells should have some reference to back up their claims. --Richlv (talk) 09:21, 29 January 2010 (UTC)

The Great Cleanup of 2010

Started on some SRS BIZNISS cleanup of this article today. Please feel free to chime in with any suggestions as we go. This thing has been getting messier and messier for way too long, and someone needs to take a scalpel to it soon or any real value left in it will be lost. Lahnfeear (talk) 16:21, 29 January 2010 (UTC)

• I've been through several major updates of this article in the past couple of days. I wanted to drop a quick note about the latest, to explain: I'm removing the "Written in" column. As far as a comparison point between two or more network monitoring systems, it amounts to largely irrelevant fluff that only clutters the table. It's an interesting factoid, to be sure, but the truth is that most applications in this field use several different technologies and languages in their architecture, and the question of "which ones" doesn't rank high enough in most people's list of concerns to warrant being featured in a comparison like this. Lahnfeear (talk) 13:48, 1 February 2010 (UTC)

• Update: PEOPLE, People, People... it's the first line of the freaking article. If the product you're adding is redlinked and can't be pointed at anything referencing it on WP, it doesn't belong here. It took too long to clean up all of the advertising in this article to let it get cluttered up with poorly-formatted and spammy entries. Start with an article on the product itself, and if it's notable, THEN add it here, not the other way around. Lahnfeear (talk) 13:54, 22 April 2010 (UTC)

just wanted to say thanks for the cleanup, it finally started to make some sense. you seem to be gone for a while now, though, and it's deteriorating again - please, come back ;) --Richlv (talk) 16:09, 10 August 2011 (UTC)

Eeep! I'm sorry, I've been neglecting it. I promise I'll give it some TLC later this week :) Lahnfeear (talk) 20:06, 10 August 2011 (UTC)

Operating system

There is no column for "operating system". Are all those tools cross-platform? Are they are for *NIX platforms only? Or do we have a mix of Linux, unix, windows and cross-platform apps? AugustinMa (talk) 11:53, 26 July 2010 (UTC)

That's a really good point. It's a mix of platforms -- I think adding an "OS" column would be very helpful to some people.Lahnfeear (talk) 14:36, 26 July 2010 (UTC)

I would also like to see an O/S column. I have Windows, Linux, and other O/S variants to monitor (we're a consulting company); takes a while to dig through all the entries that just won't work. Jgwinner (talk) 23:02, 17 March 2011 (UTC)

Open/Free vs Commercial

Would it be beneficial to split this into two articles, one of commercial and one of free/open? Generally if you're looking for information about an NMS, you're either after a free one or a commercial one, so mixing the two up just makes the article less useful. Also people looking for open systems are often looking at different criteria than those looking for commercial systems. For example, someone looking for an open system might be very interested in what language it's written in, but it's unlikely someone wanting a commercial system would care about that. Just an idea :) Adamathefrog (talk) 01:06, 2 August 2010 (UTC)

You could split into separate tables, I wouldn't recommend splitting into separate articles. --Hm2k (talk) 08:56, 2 August 2010 (UTC)

Because the table uses the sortable template, you can just click the header of the "License" column and sort everything by open source vs. commercial. Splitting it into two tables or two articles seems like a more cumbersome way of doing this. Lahnfeear (talk) 12:28, 2 August 2010 (UTC)

I'd probably agree with this if the table wasn't already getting so large... --Hm2k (talk) 13:16, 2 August 2010 (UTC)

Not trying to be combative, I'm just trying to see what vision you have for this: How does having two tables, with the same number of columns as the current table, directly on top of one another, result in the page getting smaller? I'm all about trying to make this article more usable -- you should've seen it before we cleaned it up back in January. It was GARGANTUAN. Lahnfeear (talk) 14:34, 2 August 2010 (UTC)

An article longer than one or two pages when printed should be divided into sections to ease navigation. Keep up the good work. --Hm2k (talk) 16:33, 2 August 2010 (UTC)

Part of my point was that the two types of product wouldn't necessarily have the same columns. It'd improve flexibility by having more information, and not waste space by unnecessary information? Adamathefrog (talk) 16:46, 4 August 2010 (UTC)

What columns would you add or subtract from each table? Lahnfeear (talk) 17:31, 4 August 2010 (UTC)

Host OS and License type for open source stuff instantly springs to mind, whereas they're less important on commercial (though i think i said that above). The table also slightly ignores the fact that "network monitoring systems" is a very, very wooly term, and seems to be equally applied to a range of software, from data graphing systems like cacti to up/down monitoring systems like nagios, and everything in between. Perhaps splitting it into a couple of tables based on more accurate categories would be more helpful. I'm not 100% sure :) Adamathefrog (talk) 17:23, 8 August 2010 (UTC)

In addition to adding HP OpenView, WhatsUp Gold and Dorado Redcell, it would also be nice to add a column for "Flow Analysis" which references the growing important presence of NetFlow, SFlow or JFlow collection and reporting in a number of NMS products. For those that would complain there are already too many columns, I'd recommend a larger monitor. This table is quite useful and should continue to be expanded even if that means breaking it into separate tables. —Preceding unsigned comment added by 24.229.18.194 (talk) 12:30, 23 September 2010 (UTC)

network instruments observer entries

"network instruments observer" entries look more like advertisements, not neutral data :)

• "Proprietary High Performance Database" - "High Performance" is like from marketing materials. just "Proprietary Database" would suffice.
• why the orange background ? the fact that it's using a proprietary db doesn't warrant such a position, so same background as others should be used.

if there are no objections (and nobody else changes that), i'll make the changes listed above at some point

--Richlv (talk) 15:22, 4 January 2011 (UTC)

I totally agree, it's advertisement. Didn't catch that until you pointed it out. I'll make the change. Lahnfeear (talk) 15:39, 4 January 2011 (UTC)

just wanted to say thanks for cleaning this page up - it finally makes at least some sense :) --Richlv (talk) 11:35, 12 January 2011 (UTC)

Management features in addition to Monitoring

The focus of the feature set is on Monitoring, but there is also a great value in a Management set of features. I'm sure there will be a number of products that focus on either Monitoring or Management, but it would be very useful to see the overlap in functionality of products between the Monitoring and Management.

Some features I could see others would find useful:

* Remote/Automated Provisioning
* NetConf Support

There are otehrs, but I can't think of them right at the moment. —Preceding unsigned comment added by 192.131.177.18 (talk) 14:49, 31 January 2011 (UTC)

Verbosity of Zenoss entry

Some fields of the Zenoss entry seem to be excessively verbose (in comparison with other entries) - for example, "Yes; ZenPacks". Is the "ZenPacks" part really relevant? Similar case with other columns. Looks like coming from a marketing department... —Preceding unsigned comment added by 212.93.101.208 (talk) 19:54, 9 February 2011 (UTC)

Agreed. Will edit. Thank you! Lahnfeear (talk) 20:54, 9 February 2011 (UTC)

access control

"access control" column is a bit dubious. while most entries go between "Yes" and "No", some have "Multi-tier" and "Granular". unless specific definitions for each value are provided (which i suspect might be very arguable and also make the table hard to read), i'd propose changing all those to "Yes". --Richlv (talk) 22:08, 8 August 2011 (UTC)

removed unexplained additions to "access control" column --Richlv (talk) 09:09, 11 August 2011 (UTC)

license entries

some entries for license are quite different. most of the software that is available under a commercial license but has limited free version has "Limited free, Commercial". some have it different, though, including "GPLv2; (Enterprise edition available)", "Freeware and Commercial", "Commercial (Free)" and "Freeware and Commercial". maybe it's best to unite the message and cell colouring for them ? --Richlv (talk) 11:08, 13 September 2011 (UTC)

Zabbix trend prediction

I'm pretty sure the software does not have such functionality. (http://www.zabbix.com/forum/showthread.php?t=11788) And there is nothing about this feature in the documentation. — Preceding unsigned comment added by 212.77.172.62 (talk) 16:18, 11 November 2011 (UTC)

Oops

That last revision note should have read, "...advertise Pandora FMS," not OpenNMS. Lahnfeear (talk) 14:07, 7 December 2011 (UTC)