2021 User Survey Results #6086
jeremystretch
started this conversation in
General
Replies: 2 comments
-
"Retail" is an industry that would be a sensible addition to the survey. |
Beta Was this translation helpful? Give feedback.
0 replies
-
Very insightful report. I'm surprised there wasn't more requests for DNS support, in particular. |
Beta Was this translation helpful? Give feedback.
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
-
The results from our 2021 NetBox user survey are in! I and the other maintainers want to thank everyone who took the time to participate. Your feedback helps form the direction of NetBox's development over the coming months and years. Here are the results.
Structured Responses
Similar to past years, we see a heavy skew toward information services, with healthy representation in education and government as well. However, that 20% "other" chunk suggests we might need to revisit the listed industry classifications.
Like last year, we've got a fairly even spread across organizations of all sizes, which is great to see.
For this year's survey, we decided to collect some information on how many of different types of objects each respondent has defined in their NetBox instance. These help inform some behind-the-scenes modeling decisions.
It's awesome to see so many organizations using NetBox at scale!
There's been a slight uptick in the portion of organizations deploying NetBox as a container, though the traditional VM deployment still enjoys a healthy lead. Interestingly we're also seeing slightly higher bare metal deployment.
HA deployment numbers are up considerably across the board compared to last year.
Aside from the Docker portion increasing a few points, very little change here compared to last year.
Again, not much change. Secrets usage has declined markedly and is due to be split off into a plugin for NetBox v2.12.
New on this year's survey is the plugins feature, used by nearly half of respondents.
Strong evidence that Microsoft Excel has retained its position as the world's most popular IPAM software.
The moderate upward trend across all integrations as compared to last year is encouraging.
Pretty similar responses to last year.
Open-Ended Responses
Finally, we asked: If you could add one feature to NetBox, what would it be? Free-form answers are always challenging to quantify, but here are the most popular responses (those with at least three mentions), loosely correlated:
Overlay modeling is certainly something I'd like to implement, though it's going to be a huge undertaking and will require a great deal of planning to ensure we get the models right. Firewall rule management comes up pretty frequently, but it's so broad that IMO it easily warrants its own application, especially when you consider the myriad nuances of individual vendors and platforms.
Floor plan modeling was proposed in a very early feature request but was recently rejected as a core enhancement for the same reasons we ditched early attempts at native support for network diagramming: The complexity required to develop and maintain the feature significantly detracts from the core application, and its "ideal" implementation is bound to be highly user-dependent. This sort of functionality is much better suited as a plugin, and several NetBox plugins have been created to support network topology drawings.
GraphQL could make sense as a complement to the REST API, though it hasn't been a priority. This is another great candidate for initial development as a plugin. Once mature, it could be integrated with NetBox core pretty easily.
It's difficult to tell from a few terse responses exactly what people mean when they say "ASN modeling" or "BGP AS tracking," though it's clear the area deserves more investigation. I invite anyone who has a particular model or functionality in mind to submit a detailed feature request for further discussion.
There are a lot of external authentication apps for Django, and NetBox already supports customizable authentication backends. But I'm open to specific proposals by individuals who would like to contribute a more comprehensive implementation.
We have a few minor issues open relating to power modeling, though there's nothing super exciting. It was difficult to discern from the responses exactly what change(s) were being suggested, but I encourage anyone with a specific proposal to open a feature request.
And finally: virtual circuits. Yes. At some point. Much like overlay modeling, we need to consider all reasonable use cases and come up with a robust data model that can support all of them.
In addition to the suggestions above, I want to call out the following, all of which have already been addressed in the upcoming v2.11 release:
Thanks again to everyone who participated, and keep an eye out for the v2.11 beta this week!
Beta Was this translation helpful? Give feedback.
All reactions