Talk:Simple Network Management Protocol/Archive 1

5-Ws

This is a fine technical discussion of SNMP; however; Wikipedia is an encyclopedia not a tech manual.

Any Wikipedia article should answer the 5-Ws: Who (created SNMP) What (does it do) When (was it created/revised/discontinued) Where (is it used) Why (is it used)

Esalkin (talk) 16:35, 14 November 2008 (UTC)

SNMP - No Longer Used?

Hi all, Im not to sure whats going on but according to a guy from Motorola, he claims that SNMP is no longer in operation or supported by the IETF. This was said in a talk he gave to a group of students studying various aspects of network management.

I have his email address and will ask him to email me back with some proof of that, but he assured us in the talk that he was not liying - and I can guarntee you many people in the room who lectured in that area were pretty supprised.

Anyway, just thought id give the heads up.

--Sully 22:03, 19 April 2007 (UTC)

First off, I'm not overly interested in what the IETF thinks as SNMP is Mission Critical where I work. We also know Motorola as a very poor implementor of SNMP. They much prefer to use their own proprietary NMS tools. I can't speak for the Comcasts of the world, but the mid-sized MSO I work at uses SNMP very extensively. marnues 23:20, 29 October 2007 (UTC)

marnues is right of course - it's what people are using that really counts. But to clarify the apparent IETF position, I have edited the article slightly. Internet Standard (STD) is as serious as RFCs get, and it's rare that they get this far (most never make it past "proposed standard"). Anything they no longer think relevant is marked "HISTORIC" (and they do have an active program of marking old stuff this way). That they mark it "STANDARD" says they think it's finished (as in complete, not out of date). Not that everyone agrees with the IETF view of course, or that it doesn't have its problems, but the fact that it's also in widespread active service indicates that it's certainly not a dead technology. Behind The Wall Of Sleep 19:59, 9 November 2007 (UTC)

If that were true, they wouldn't keep publishing new MIBs. But they do. There are new technologies emerging and users will always have choices to pick from. Use what is right for you based on availability and need, rather than marketing hype toward or against SNMP Hardaker 21:06, 9 November 2007 (UTC)

Compatibility between versions?

Can someone please add some data regarding the extent of (backward) compatibility between SNMP versions?

For instance, can devices using SNMP v3, v2(whatever) and v1 operate efficiently on the same network . . . ignoring security issues (in the event it is a standalone environment)?

---

I am a little surprised that there is no discussion page for the SNMP Article. There are always lively discussions at the IETF meetings. Wes Hardaker, where are you? Randy Preshun? David Levi? Anyone out there? David Battle

Who wants to deal with SNMP more than they absolutely have to? Ha ha, only serious. Then again, Vint Cerf keeps on mentioning once ever few years about how he's going to write SXXP or BXXP or whatever the successor is going to be.... --moof 04:25, 11 November 2005 (UTC)

---

I added the 'section cleanup' tag to the 'Architecture' section. The section starts by listing the three components of an SNMP architecture, but fails to define either of the first two. It makes offhand reference to the second, but does not even mention the first.

I am only now learning about SNMP, so the lack of those definitions makes the page nearly useless. Could someone with a greater knowledge of SNMP fix this?

Maplebed 18:19, 28 September 2005 (UTC)

It's been a while since I did serious work with SNMP, but I've tried to fill in the gaps although the article remains rough around the edges.

Ray Dassen 19:28, 28 September 2005 (UTC)

Hi, I'm new to editing in Wikipedia. Currently I am studying about SNMPs too. And I have made a few changes in the Architecture and MIBs. I have also added a new section about Proxy agents. Please review and make changes as necessary.

Yes, but what does it do?

I notice that this article says a lot about the history and the structure of SNMP, but very little about what it actually does or why. Can someone more knowledgeable than me help with this?Urocyon 04:12, 13 March 2006 (UTC)

I've added a bit about router graphing software (which builds on data gathered through SNMP) to the example section. Hope that helps somewhat. Ray Dassen 06:26, 13 March 2006 (UTC)

I fleshed out some of the possible uses, with a few starter points to get new people off and running. I also noted a few of the major gotchas that come along with SNMP deployment. Billndotnet 00:06, 18 June 2006 (UTC)

I added both a refference to the Network management model that defines the reason for snmp and to FCAPS that defines the functionality of snmp--Mancini 16:07, 26 November 2006 (UTC)

This is a great technical review that anyone already well versed in SNMP will have no problem following. It is useless in Wikipedia! This article should be written in layman's terms. Move the current article to a glossy networking magazine where it belongs. -- esalkin 12:49, 12 June 2007 (UTC)

This seems to be plagiarism.

The introductory paragraph is nearly identical to the Cisco resource found here:

http://www.cisco.com/univercd/cc/td/doc/cisintwk/ito_doc/snmp.htm#wp1020581 —The preceding unsigned comment was added by 64.236.240.190 (talk • contribs) 03:34, 3 December 2006 (UTC).

RE: SNMP is no longer in operation

Thanks for the heads up, Sully.

It may have been dropped by the IETF (I don't know) but it still very much in use in the real world.

It is the most widely used management protocol of its type. I don't have references for this but I have worked in the networking industry for over 10 years and this is my experience.

Endpoint 11:21, 5 May 2007 (UTC)

Proposal to move UDP paragraph into separate section

If there are no serious objections, I would like to move the section on UDP ports into a separate section, named something like 'SNMP at the transport layer'

At the moment it is in the section 'SNMPv2 and Structure of Management Information' but it is really about how SNMP uses the transport layer.

UDP ports section reproduced below:

"Typically, SNMP uses UDP ports 161 for the agent and 162 for the manager. The Manager may send Requests from any available ports (source port) to port 161 in the agent (destination port). The agent response will be given back to the source port. The Manager will receive traps on port 162. The agent may generate traps from any available port.

Many distributions change this, however, and this is not necessarily always true."

--Endpoint 11:26, 5 May 2007 (UTC)

I agree with this. - 194.237.142.6 (talk) 13:23, 5 September 2008 (UTC)

orphaned pages

Network status gathering system seems to be related and is a orphaned article. —Preceding unsigned comment added by 76.112.118.208 (talk) 19:29, 7 September 2007 (UTC)

Really bad description of SNMP

This Wiki page is a really bad description of SNMP. It misses in identifying and describing the fundamental concepts of SNMP.

-- an SNMP RFC editor and co-author, and long time IETF participant

It sounds like you would be a great person to help fix it! Which fundamental concepts would you like to see addressed? Lukeritchie (talk) 23:56, 28 February 2008 (UTC)

Question about parsing of the text

In the sentence "A managed object is a characteristic of something that can be managed", is it the characteristic or the thing that is managed? —Preceding unsigned comment added by 209.234.91.82 (talk) 20:14, 13 March 2009 (UTC)

Plagiarism

This article IS NOT plagiarism. The original can be found here: Cisco SNMP article, which is copyrighted. —The preceding unsigned comment was added by Darkmane (talk • contribs) 10:39, 31 January 2007 (UTC).

---

I tend to disagree.. I took a look at the cisco article offical, and this one here and did some comparing. The content under "Management Information Base (MIBs)" is very similar to that on the offical document. Its like a copy and paste job with a clean up style wise.

Parts of this article may be considered plagiarism, but not the whole anyway.

--Sully 22:09, 19 April 2007 (UTC)

---

There are several unusual phrases that can be googled and they hit 2 places, this article and the Cisco paper. Try "SNMPv1 is widely used and is the de facto network-management protocol in the Internet community" for example. This article is plagiarism in part and badly written overall. —Preceding unsigned comment added by 67.79.13.14 (talk) 18:28, 12 January 2010 (UTC)

Yes, but how does it work? What does it look like?

Is it a binary protocol? Or a text protocol? Or vendor specified? What does the packet/message/whatever look like? The description is very abstract - the article says that there are some types and variables and the ASN describes only the high level structure but not the implementation. The RFCs are great - as a reference for people who already know how it works. I went quickly throught the first listed and there is still only the high level structure. —Preceding unsigned comment added by 80.95.97.145 (talk) 08:02, 1 December 2009 (UTC)

Realizing this might be better in the text than here, but none the less: the protocol is encoded in BER which is a binary encoding. —Preceding unsigned comment added by Hardaker (talk • contribs) 15:31, 1 December 2009 (UTC)

Yes, but how does it work? What does it look like?

Is it a binary protocol? Or a text protocol? Or vendor specified? What does the packet/message/whatever look like? The description is very abstract - the article says that there are some types and variables and the ASN describes only the high level structure but not the implementation. The RFCs are great - as a reference for people who already know how it works. I went quickly throught the first listed and there is still only the high level structure. —Preceding unsigned comment added by 80.95.97.145 (talk) 08:02, 1 December 2009 (UTC)

Realizing this might be better in the text than here, but none the less: the protocol is encoded in BER which is a binary encoding. —Preceding unsigned comment added by Hardaker (talk • contribs) 15:31, 1 December 2009 (UTC)

Varbind missing

The term "Varbind" is redirected to this page. This page, however, does not contain the word "Varbind." —Preceding unsigned comment added by 206.169.32.3 (talk) 19:24, 27 August 2009 (UTC)

Inclusion of reference to xkcd as explanation for entropy

I question referecence 12, i.e. the reference to an xkcd comic, for demonstrating password entropy and perceived usability of long passwords. If it is deemed desirable that such an explanation be included, a reference to a research paper should be included. However, I do not think that explaining useful passwords is an important contribution to an article about SNMP, either. 131.159.20.131 (talk) 14:33, 17 August 2011 (UTC)

Then be bold and revert it. BollyJeff || talk 14:46, 17 August 2011 (UTC)

Indefinite article

See Wikipedia:Administrators' noticeboard/Incidents#User: 109.77.xx.xx and the indefinite article and Talk:XMPP#Please discuss changes to the indefinite article. Andrewa (talk) 15:08, 24 May 2013 (UTC)

A list of SNMP Features would be useful

To help the newbie understand the 'why' of SNMP, it might be useful to start this article with a list of SNMP's features, such as

• provides a way to ask a device on a network to describe itself.
• information is organized by keyword and a corresponding value, for example you can ask a device for its 'sysLocation' and receive a response like 'Room 202A'.
• There is a single family tree (hierarchy) of keywords for all of SNMP, which makes it easy to attach new branches to this tree as new needs arise.

-- The preceding unsigned comment was added by someone whom the unsigned comment bot failed to catch.

I'd say more explanation than that is due, but your comment certainly explains more about it than the entire article. Usually the first paragraph of an article gives you a good general overview of the topic, but the first paragraph amounts to little more than what the first sentence alone says: Simple Network Management Protocol (SNMP) is an "Internet-standard protocol for managing devices on IP networks." The quotation marks around the definition there are a nice touch: it's as if it's some sort of acknowledgement that even the author couldn't pin down an actual explanation of what the protocol does and didn't want to present such a vague definition as their own work.

So I continued to read the article, but never found an answer about what exactly the protocol does. It just keeps referring to what SNMP does as "management" without ever explaining what is meant by "management." So I'm left wondering, what does "managing devices" mean? I sometimes manage my network devices by unplugging cables, moving them to a different room, and plugging in new cables. Does SNMP do this for me? Sometimes I manage my network devices by assigning them IP addresses, does SNMP do that? Sometimes I manage devices by upgrading their firmware. Does SNMP do that? Sometimes I manage my router by adding new port forwarding rules. Does SNMP do that? "Manage devices" is a uselessly vague definition of what the protocol does, whatever that might be.

My only guess is that the protocol doesn't actually do anything, but rather, it's somewhat like TCP and UDP wherein you might use it to transfer information between devices, but exactly what data any device chooses to send, receive, or act upon, is entirely undefined, and depends on the software running on those devices. The only difference being that maybe it defines a standard format for that data, like "variable name goes here" and "an ID code for the data type goes here" and "the data goes here," but otherwise the protocol doesn't attempt to define what data should be sent or what actions should be taken in response to it. ...but, I really have no idea. The entire rest of the internet seems to be equally clueless about what this protocol actually does and what it might be used for and so I've been unable to figure anything out.

-- The one and only Pj (talk) 19:44, 24 October 2013 (UTC)

Addendum to Security Implications

There have been some additional security issues with SNMP that have come up lately, and should probably be listed. The vulnerability mentioned is related to autodiscovery, so it might be appropriate to make mention of it there as well. Link to the original paper: https://www.usenix.org/system/files/conference/woot12/woot12-final14.pdf — Preceding unsigned comment added by 68.107.99.224 (talk) 20:00, 27 October 2013 (UTC)

SNMPv2 is not a Draft

IMHO: SNMPv2c's RFC 1901 is not a "Draft Standard". It is at the experimental track and now is declared historic. The Draft Standard RFC 1908 has simply involved into RFC 3584.

OK, we can check the status of v2 docs here: at the RFC Editor. A handful of RFCs were designated Historic in 2002, but we should bear in mind that this was a specific action and isn't particularly common practise. Historic sends a fairly strong message (that the document is "not an Internet Standard of any kind", see [1]) and tends to happen when such a statement is intended. We can see that 1441, 1451 and 1901 are Historic. 1442 - 1444, 1448 - 1450, 1452 are Proposed Standards but have been obsoleted. 1902-1908 are Draft Standards which have been obsoleted. This sounds a bit incongruous but it is the intent of the IETF and we should reflect it: eg they have the maturity of DS but a newer, better standard exists (v3, which is a full STD).

Also we have an outstanding citation needed for about four years on "SNMPv1.5". looking at [2] (1992) and [3] (1994), the earliest drafts in each case, both are clearly called "SNMPv2". So I think it reasonable to lose this reference unless we can find an authoritative source for "v1.5" (say notes from a WG meeting or a draft by one of the authors). Behind The Wall Of Sleep (talk) 15:34, 24 February 2014 (UTC)

Internet Culture & History - Simplicity & Running Code

This article deals with the simplicity and absent features (missing complexity) of the SNMP remote management protocol. This question (Why missing?) is an entry point for a missing treatment of the deeper culture and history of this protocol, of the prime mover in its creation Marshall Rose, and of the Internet itself, which these people and that culture created fast and first among many attempts to move global telecommunications from circuit-switched to packet-switched network technology. Rose was also a prime mover in the creation of the Plain Old Postoffice (POP; Post Office Protocol) protocol used by millions to collect their mail.

Here is some inspiration to authors better qualified than I (steal whatever you want, this is Wikiland, plagiarism is free). From "How Anarchy Works -- On location with the masters of the metaverse, the Internet Engineering Task Force." by Paulina Borsook in Wired (quote) MIT professor Dave Clark, one of the grand old men of the Internet, may have unintentionally written the IETF anthem in his A Cloudy Crystal Ball/Apocalypse Now presentation at the 24th annual July 1992 IETF conference. Today, it's immortalized on T-shirts: "We reject: kings, presidents, and voting. We believe in: rough consensus and running code." Which might translate to, "In the IETF, we don't allow caucusing, lobbying, and charismatic leaders to chart our path, but when something out on the Net really seems to work and makes sense to most of us, that's the path we'll adopt." (/quote)

After establishing that it was intentional and historically characteristic that the SIMPLE Network Management Protocol be simple, this article can grow and go on to enumerate the development since the 1990s that have strained that simplicity and widening application of the SNMP: cheap memory, ever-larger firmware, protocols with ever more options, very long logs and tables (e.g., router tables, switch tables) that a larger network requires and larger memories can store, and a public and global network that now embraces bad actors.

Much can be taught to many from this little SNMP article. The SNMP is a small, concrete piece of the Internet and its history. We will surprise readers with how much makes sense and is learned. Jerry-VA (talk) —Preceding undated comment added 19:41, 17 September 2015 (UTC)