Jump to content

Talk:DNP3

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

DNP is as far as I know only used in elecric and water utilities and only in USA and Australia. Is it actually still used? or is it a vintage protocol like 870-5-101?? HaPi

DNP3's use is on the rise at least here in New Zealand. The company I'm working for is upgrading all the RTU communications to DNP3 from conitel.
The US and Australian SCADA markets are not small. Furthermore, the WITS Consortium in the UK has standardized on DNP. DNP growth may seem threatened by protocols such as IEC-61850, though they're really complementary in most respects. Scadateer (talk) 15:16, 4 July 2008 (UTC)[reply]

Technical Detail Edit

[edit]

I have updated the Technical detail section with more accurate and more useful overview of the protocol. Full disclosure: I am a member of the DNP Technical Committee. While I not exactly an unbiased, technically knowledgeable spectator, I did try to maintain a sense of perspective.Scadateer (talk) 18:56, 4 December 2008 (UTC)[reply]

Added IEEE standards update for 2012 and a very small overview of differences between these versions. Scadateer (talk) 21:27, 16 April 2013 (UTC)[reply]

DNP Vs Modbus

[edit]

Can I Know How Different is DNP from Modbus as far as Data aquasition is conserned?


DNP is a great SCADA protocol. Modbus is a good DCS protocol, but not a good SCADA protocol. It wasn't designed with the supervisory features that DNP has.

First, Modbus has no concept of event based reporting. In other words, you have to ask it what everything is each time. There is no way to ask "what's new?" in Modbus. This greatly increases bandwidth consumption. On low-bandwidth or expensive bandwidth links such as a VSAT terminal, this is a very big deal. In contrast, DNP can report events such as an analog value that has gone outside its deadband.

Second, DNP has a much tighter object library than Modbus. Modbus systems often pack data in to various registers. This is unnecessary with DNP. Furthermore, Modbus is not well defined for issues such as Endian-ness. There is no standard way of sending the time in Modbus. There are no floating point number representations in Modbus. These are common features in DNP.

Third, DNP now has secure authentication features to protect against a man-in-the-middle attack. The communications might be denied, but they can't be spoofed in to doing things that require authentication. So, no switching large feeders in substations without an authenticated packet. The only option for Modbus is encryption, as in AGA-12. The bandwidth overhead for encryption is usually higher than it is for authentication. Scadateer (talk) 15:05, 4 July 2008 (UTC)[reply]

Minor Issues

[edit]

The Security section has a hanging "g" character. The IEEE 1379-2000 specification has been withdrawn as of 6 Dec 2011. The hyperlink points nowhere. Perhaps because of the passage of 1815 which is mentioned later. Davesnotthere (talk) 15:04, 29 June 2012 (UTC)[reply]

Advertising

[edit]

Where is it reasonable to draw the line for advertising? The Training section is an example of my question. I removed the Subnet reference for training because I have confirmed that they no longer offer this service; but that still begs the question of whether it is appropriate for Wikipedia standards.

Also, the open source implementations are not being offered independently from the companies that wrote and consult with them. Is this regarded as free advertising? Does Wikipedia have a policy regarding these implementations, or should they stand as is? (Jacob Brodsky; Chairman DNP User Group) JakeBrodskyPE (talk) 20:20, 9 July 2013 (UTC)[reply]

Why was opendnp3 removed in commit 729615371 ? It is a complete implementation of the protocol and it is very useful for someone learning about it. Also, the decoder on their website (which was in the external links section) helps understanding the topic by providing detailed explanation of a DNP3 transmission. Also, the link to Triangle Microworks with the training videos was relevant. The videos are still there and freely available. They represent a good primer to understand the protocol. I would recommend reverting commit 729615371. — Preceding unsigned comment added by Emgre (talkcontribs) 22:53, 15 July 2016 (UTC)[reply]

The external link policy for Wikipedia is covered here. The editor has cited it, but has failed to cite which specific policy the links and content violated. In full disclosure, I am the principal maintainer of opendnp3. I occasionally consult on opendnp3 and DNP3 related topics.

Going through the policy points in order:

  • 1) N/A - All removed links contain substantial relevant content beyond the article.
  • 2) N/A - Content is technical in nature, not based on opinions.
  • 3-> 5) N/A
  • 4) N/A
  • 5) N/A - The primary purpose of the opendnp3 link and content is to provide a free working library to anyone, and a free forum for asking questions about DNP3 in general. The vast majority (>95%) of opendnp3 users/companies do not pay for any services, and the IP itself is the copyright of > 10 organizations. In theory, anyone could consult or sell solutions based on this IP since it is Apache licensed. Based on this reasoning, the primary purpose of the link is to benefit the DNP3 community, not the contributors to the library or anyone who might consult on the project.
  • 6) N/A - None of the content from the removed links is behind any kind of paywall.
  • 7 - 13) Clearly N/A
  • 14) N/A - Open source projects are NOT "manufacturers, suppliers or customers". The videos from TMW, while they are a commericial supplier, are primarily for educational purposes not promotion as well.
  • 15-19) Also clearly N/A

The specific content regarding the security research related to DNP3 and opendnp3 was covered in well-respected industry blogs like Digitalbond in addition to the mainstream media like the New York Times and Wired Magazine. This is clearly of importance, but I welcome anyone else to make the language more neutral if they see anything out of sorts. I have a conflict of interest being the principal maintainer of opendnp3, so I will let someone else revert commit 729615371 or modify the content to make it suitable. Jadamcrain (talk) 21:34, 16 July 2016 (UTC)[reply]

The IEEE-1815 standards are a difficult read, especially those new to the protocol, and resources such as the opendnp3 source code the training videos from TMW are crucial for fully understanding how the protocol works and is actually used in industry. (Full disclosure: I have contributed to the opendnp3 project, although I am not a principal maintainer). sidhoda —Preceding undated comment added 08:14, 18 July 2016 (UTC)[reply]

+1 The free content that was removed is valuable for understanding DNP3 and should be restored. I'm a user of the opendnp3 and had occasionally contributed in the documentation. I never paid for any commercial services. Opendnp3 is an excellent tool no learn about the protocol and to test different configurations. pgiudice comment added 17:06, 19 July 2016

Considering the fact that many people agree that the links were relevant, and considering that, to my knowledge, no Wikipedia policy was violated, I have reverted the changes. I am affiliated neither with Automatak nor Triangle MicroWorks. I am just someone learning DNP3 and using opendnp3 as a learning tool. Emgre (talk) 22:51, 20 July 2016 (UTC)[reply]

Videos fall afoul of WP:ELNO points 7 and 8. Links to implementations are off-mission for Wikipedia - in particular, Wikipedia is not a link directory. This article is a place for laypeople to read about what DNP3 is, not a place for developers to find training materials or to list or find software that implements the protocol
Also, welcome to all the new editors, but everyone here should know that using off-site notices such as [1] is not looked on very positively on Wikipedia, see WP:CANVASS. - MrOllie (talk)
You have not established that the content is not relevant to the topic of DNP3, nor tied it to a particular type of "link to be avoided". Instead you have now devolved to a very generic argument "Wikipedia is not a link directory". Since this is very vague and nonconstructive, it is difficult to argue. In waht way does this turn the page into a link directory?
For consistency sake, I now invite you to edit the many hundreds of protocol/specification pages on Wikipedia that link to open source reference implementations. These three should get you started:
Advanced_Message_Queuing_Protocol#Implementations
MQTT#MQTT_broker
Secure_Shell
After you have completed the task of removing all links to open source protocol implementations from these many hundreds of pages, I might be persuaded that you are engaging in anything other than stubborn bullying.
Seriously, there are no dissenting opinions here from anyone knowledgeable about DNP3, and input has been voiced by unaffiliated 3rd parties. — Preceding unsigned comment added by Jadamcrain (talkcontribs) 01:32, 21 July 2016 (UTC)[reply]
Indeed, I do spend quite a lot of time pruning external links from various places where linkfarms are prone to accumulate. Thanks for those 3 examples, some look bad, but I don't really see what you're getting at with Secure Shell. I'll add them to my list. But Wikipedia is a big site and there is a lot to do here - that you have found some articles with inappropriate links is not any kind of reason to add more such links to this page. To answer your specific question, that link turns this page into a link directory because Wikipedia is not in the habit of linking to suppliers (ELNO point 14), and when one such link is permitted inevitably more get added. I happened to notice this article because it was one of several hit by a user who was repetitively adding links to his company to several such places on Wikipedia. To be frank, I don't really see a lot of distinction between the addition of that link and your additions of automatak. PS: I see that the offsite canvass has been deleted, thanks for that. - MrOllie (talk) 01:47, 21 July 2016 (UTC)[reply]
As has already been argued above, open source projects, that anyone can consume for free are not suppliers. "Supplier" implies that you can't get something without paying for it. This isn't what open source is about. Trying to cite ELNO point 14 is quite the stretch. — Preceding unsigned comment added by Jadamcrain (talkcontribs) 02:01, 21 July 2016 (UTC)[reply]
This is one way that Wikipedia and the open source movement differ: Wikipedia draws no distinction between for-profit and non-profit parties. To quote WP:EL: 'Wikipedia uses the same standards for evaluating links to websites owned by for-profit and (real or purported) non-profit organizations.' - MrOllie (talk) 02:08, 21 July 2016 (UTC)[reply]