I would submit some ideas for NethServer, not for features but about our communication in general:
-Improve communication for a better divulgation.
For exemple, go to Distrowatch, and simply seach for a Firewall distribution…
Nethserver not appear in the search result.
So I suggest to define better what is Nethserver, and hilight advantages and at the same time take care to distinguish Nethserver from concurence.
When this step is done, we will ready to make a good “marketing” campain with Distrowatch, Zdnet, Techradar, and other medias.
To start:
Is the “all in one” definition is good? Or it better to hilight the modularity, the versatility?
The all-in-one concept suggest a sort of melpot, that Netserver doing all at the same time. Something monilitic, a big brick.
When the modularity is suggesting “feet exactly your need”.
In place of all in one, I would suggest to define better what Nethserver is, and what Netserver can do: Firewall Gateway / Ad Server / Hypervisor / web server / file server
Interesting point indeed, as @GG_jr already said “what NethServer should be” is a good question.
I think that we have to focus on what NethServer is different from other similar distributions/product
As for as I know, newcomers and people from the outside said multiple times:
nice and active community
all-in-one BUT modular and customizable
CentOS complaint
many modules pre-configured and point-to-click
a modern, fast and easy web interface
I agree,we have to improve communication for a better divulgation
I like the idea to refine the “Catch phrase” for NethServer and also to better define it within the Server Windows/Linux world.
As @Ctek said me once: “We must have a brainstorm and see what can be refined and what can be re-invented, defined that will “explain” NS to the reader and possible interested users.”
@robb what do you think about ‘all-in-one’ solution?
Should we change our definition?
NethServer is an operating system for Linux enthusiasts, designed for small offices and medium enterprises
One way to find our communication is to define our place in all the existing Distros, of course eliminating all desktop distros.
Pfsense, Opnsense, ipfire, Endian, Untangle, Sophos and others Firewall/Gateway distributions.
But NethServer do more…
There’s ClearOS, SME Server, Zentyal, Yunohost, Univention, Amahi, perhaps TurnKey Linux, If there’s other distros, please tell us.
Between these distros, in what NethServer is different?
When we will have this answer, it will be easier to define NethServer. It will be easier to hilight our strenght.
Personnally, I bet on :
Modularity and versatillity. For exemple, defining what NS can do and what it can’t do yet.
Free, Open Source, and community driven. it’s community driven, but what we can do, and what we can’t ( I think about law, copyright, why other community have fundation, and so on… )
Clean Webgui.
But as always, I like debates, I like to know what you think
I already said before what I think about “all-in-one” server. I think this should be one of the main arguments, but at least as important is the scalability, flexibility and the modularity of NethServer.
I agree that on sites as Distrowatch.com, Alternativeto.net etc, all the different services should be mentioned, and not just “all-in-one” or “Small Business Server”. People get to these sites because they search on UTM or Gateway, Mailserver, Fileserver or Printserver. Then these keywords should be defined in the NethServer profile of that site so NethServer pops up when such a keyword has been entered.
To add to @Jim’s list, I will add my ‘other distro’ that is the complementary server of the GW/UTM distro’s: Karoshi Linux. It is all (plus a lot of educational services) NethServer can do but lacks a GW/UTM.
“Improve our communication”
or
“What NethServer should be”
or
“NethServer vs …”
or
“Is NethServer better than a dedicated utm distro?”
A lot of ideas and discussions.
Let’s talk on a real case: A small company which offer IT&C services to other companies from public sector and private. The goal is to replace all Windows Servers, when is possible, with Linux Servers, free distributions.
The network infrastructure is shown in the below picture (in this moment there are no NS2, ownCloud, BDC/AD and NS1 is in the RED (WAN) zone).
The Internet access (inbound/outbound):
a public IP subnet routed through two ISP with BGP service (a primary connection and a secondary/backup connection).
The servers:
UTM: Endian 3.0.5 beta 1. Can be replaced with other distro? Why?
NS1: Authoritative Name Server - Win 2003 (will be replaced with Bind on Ubuntu I think) - Can other distro be used? Why?
NS2: Secondary Name Server - will be Bind on Ubuntu I think - Can other distro be used? Why?
EMAIL HOSTING: Zentyal 4.2 - Can be replaced with other distro? Why?
WEB HOSTING LAMP: Ubuntu, WP - Can be replaced with other distro? Why?
ownCloud: will be NethServer - Can other distro be used? Why?
osTicket: Ubuntu - Can other distro be used? Why?
PDC / AD: Windows 2008 Std. Will be replaced with Zentyal 4.2 - Can be replaced with other distro? Why?
Video Surveillance: Windows 2008 Std. Will be replaced with NetServer & ZoneMinder ? - Can be replaced with other distro? Why?
Can be this example a good subject for future productive discussions?
Can this example give us proper responses to our questions: “What NethServer should be”, “NethServer vs …”, “Is NethServer better than a dedicated utm distro?”, Web GUI or CLI?. …
@GG_jr, tbh I missed the clue completely you are trying to point out here. Do you mean NethServer when you say “other distro”?
2nd: Why would a company choose to use so many different flavors of OS? This will make maintenance so much more complicated and will demand so much more of the admins.
If you want to get answers on IF NethServer could be a reliable replacement for each dedicated server, I think it is a very valid question. Then we also would know if NethServer would pass the test of being better or at least equal to the current situation.
It also would give direction on how NethServer should evolve if it wants to the better solution in every given service.
Why do we leave out certain features? Why do we implement certain features? If there are multiple solutions for implementing a feature, why do we choose a certain feature over another. These basic choices should be very clear. These choices ultimately will define the roadmap for NethServer.
You’re right! My mistake. Should be: Can be replaced with NethServer or other distro?
The question “Why” is for define why is a distro better than other. Why is NS better than others. I think that if we understand “why” we will know “why”?
There are only three OS: RH for Endian (everything is from GUI), Windows and Ubuntu for the others.
Like I said: The goal is to replace all Windows Servers, when is possible, with Linux Servers, free distributions. Can be used only one distro? Which is that? Is NS that distro? I will be glad to be!
I think you have this as a real life usecase? If that is the case, you could easily make a benchmark for each service. And I think most of the services can be replaced with NethServer, but in this usecase, you do have Samba4 available and that is still lacking on NethServer. And that will be a huge disadvantage and probably a showstopper in this particular case.
The good news is that NS7 will get Samba4 and multi site. So I guess in your situation you will have to wait to propose NethServer as a full replacement for all the servers until these features are available.
In the meantime you can start with the benchmarks!
Yes, from about two years.
If I should replace something, I want to have the proper answer.
My opinion maybe is subjective. That’s why I tried to give a real example.
The Email server has 9 different domains with a total of 200 mailboxes.
The UTM make all tasks: Web Proxy (HTTP and HTTPS), SMTP Proxy, POP3 Proxy, antivirus, anti spam, web filter, multi wan, Firewall, DNAT, VPN, IPS, Logs and Reports, …
On both servers everything is done only from GUI. No CLI (in two years I use CLI for max. 20 times).
I wrote all this (not only in this post) trying to make a lot of community members to understand that if we want to make NS user friendly and the proper replacement for other distros this is the right way. Linux became widely used (for non pro) since has friendly GUI. You know that you can run Windows only from CLI. But who will use Windows in this case?
Maybe I’m not so coherent. I want to tell more in short time.
IMHO there are a lot of endless discussions about nonessentials things and other essential are completed quickly: only CLI, GUI never, … Why others can do?
.
Maybe I want / ask to many and I should not to do that. I’m not a programmer and if I can’t do, why I ask?
Maybe I am too old and I can’t understand some things.
no No NOOO… keep them coming! Comments and questions like yours are VITAL to get a clear view on what direction the project should go. Keeping the discussion active is important.
I hope more members join these discussions so we have more information about what is important to the community and what we should focus on.
More on focussed on your situation:
multi domain support for email? Does it exist for NS and is it a true multi domain support or 1 domain and several virtual domain that are connected to the first domain? (still not knowing enough about NS… )
UTM of NS is extensive enough for your case? If not, what is lacking?
As soon you identified features you need: make a feature request! Explain why you need the feature and how you solve it now. add points on how you think (even if you are not a dev) the feature could be implemented, or if there are multiple options, give more options. Then we can all discuss about the feature and the possible solution.
I know you are not doing this but just dropping a line like: “I want THIS FEATURE (and I want it now)” is not helping the project. Explanation and options (some self investigation) is very important to make it possible, or even giving a reason, for others to chip in with ideas, opinions and eventually, code.
Is not about only my situation.
I give this example because is real and is complex enough. Let say it’s a case study for business application of the NethServer. I always thought to NethServer as a business product and then as a home product, not viceversa.
Many of discussions are about how to install a lot of applications which are nothing to do or related with NethServer as product. Those apps can be installed on CentOS, on Ubuntu, …
I always ask for apps as integrated modules that can make NethServer a competitor to Endian, Zentyal, iRedMail, …, and also a very good tool for administrators.
Also, IMO, if you want to beat the competition, you must know their products.
The designers of the product must know very well the competition.
Let’s talk about the UTM as the core for this network (IMO, for any network) with an example.
If we look to the DMZ, we will see a lot of servers which use HTTP and/or HTTPS protocols.
I am very lucky because I own a private IP subnet. Usually, the ISP give no more than 3 private IPs. So, what do you do in this case? You will need another server for reverse proxy. For me, the competitors for NS as UTM are: Endian and Sophos. No one of them have reverse proxy for free versions, only for paied versions. Why can’t we have Reverse Proxy on NethServer as integrated module? Of course with GUI. I bet that the answer will be: for a few lines with CLI it’s not worth it.
Let’s talk about the Email Server.
I use it for different domains: company 1.ro, company 2.ro, …
IMO, the competitors for NethServer are Zentyal and iRedMail.
For this approach, iRedMail should be the best solution because has completely separate domains for administration and for sharing contacts and calendar separately for each domain. But you must paid for full admin functions.
Because neither Zentyal or NethServer are not designed for work with separated domains, I ask here: Sorting Users and Email addresses by "Domain" for a minor thing. Nothing.
I gave some example from Zentyal about email log files. Nothing.
I like the idea to make NethServer modular but i know that is very complex design and implementation.
Look at the CUPS. I configured NethServer as UTM. Nothing to do with Print service. In /var/log/messages hundreds of lines like this: Jan 10 03:38:37 ns-utm smbd[10546]: Unable to connect to CUPS server localhost:631 - Connection refused Jan 10 03:38:37 ns-utm smbd[30880]: [2016/01/10 03:38:37.736297, 0] printing/print_cups.c:528(cups_async_callback) Jan 10 03:38:37 ns-utm smbd[30880]: failed to retrieve printer list: NT_STATUS_UNSUCCESSFUL Jan 10 03:51:38 ns-utm smbd[10646]: [2016/01/10 03:51:38.376735, 0] printing/print_cups.c:151(cups_connect) Jan 10 03:51:38 ns-utm smbd[10646]: Unable to connect to CUPS server localhost:631 - Connection refused Jan 10 03:51:38 ns-utm smbd[30880]: [2016/01/10 03:51:38.377682, 0] printing/print_cups.c:528(cups_async_callback) Jan 10 03:51:38 ns-utm smbd[30880]: failed to retrieve printer list: NT_STATUS_UNSUCCESSFUL
And is not only about lines. It’s about traffic and the system works for nothing.
Another thing: if you want to see if there are any updates, you must go to the Software Center section. After you go there and the system has finished to check available packages, you will see a message for updates if are available. If I remember well, when I have installed first time NethServer (I think was 6.5), after login, in Dashboard was a message for updates, when the updates was available. Or I confused with Zentyal.
And IPS. From my point of view, the module doesn’t work because there is nothing in log file. How can I do something against nothing?
And we can continue.
Maybe there are minor things that don’t worth attention. But can make the difference.
Again, I thought it can be a case study for all to design, make, implement a very good tool for business environment: NethServer.
I think you and I are absolutely on the same path! And I believe a lot other members of the community think like this too. If it were to me, this kind of thinking (analyse what the competition is doing and do it better) is key to make NS a better product.
That is also the reason why I encourage new members to ask questions and if there is no (satisfying) answer, ask again, and add more info. Keep the discussion going! A good way can be to post a new discussion as a linked discussion of the original discussion.
The community is less than 1 year old and both NS itself as the community change every day. New ideas arise, new features will be implemented, new options are available.
Just imagine what will happen when Samba4 is implemented… it will mean a HUGE jump in possibilities for NS and the community… Old requests that couldn’t (easily) be implemented will be possible much easier… So, repost your question, and again, keep the discussion going!
I’d like to share my experience as a consultant selling support services to SMEs.
When I started, nearly 20 years ago (I was much younger), I used a different product at each new customer.
I choose the product that, feature-wise, was more close the requests of the customer, then I configured it and sometimes even slightly modified it.
But I also used different products as an opportunity to learn something new, comparing pros and cons of each solution, making the same thing in different ways, etc.
Then, when I had more than a handful of customers, I found that it was impossible to know all those products very well and I could offer a better service focusing on knowing perfectly a single product or two.
So I selected what I liked most and started using only it, sometimes adding missing features, or improving weak spots.
I still work on consolidation of many services on a low number of systems, SMEs can’t afford to maintain a lot of server and I do a better job when I have to deal with fewer systems.
That said, my customers needs are probably different from those of other’s, and I focus on mine.
But this discussion is really important and interesting. Maybe it’s a different topic, though.
I will be always behind it, getting everyone on the linux administration by CLI is hard, this is a steep learning curve.
But keep the product balanced between a bloat-ware and " stuff only by CLI" is hard too, so we need to be very good at this sort of thing.
For me, our communication should emphasize on all-in-one BUT modular/customizable plus a modern, fast and easy web interface
)
