PDF Tutorial with Best Practice Examples? (Tell me why I'm stupid)

[Luminousdolphin]

I've bit the bullet and went for the FreeNas Mini rather than a home build. Priced out about the same, and after reading cyberj0cks review I was impressed.

So, after getting the nas and disk and setting it up (in a very basic way) and a bit of getting my feet wet, reading various forum posts and external blogs on the topic, my head is spinning. Mostly, it seems that whenever one looks for a "guide" of some sort, or a step by step, half of the dialogue is "well, this isn't best practice, but here do this" or "there's no one way to do this" or a lot of conflicting information / opinions. This seems especially so when some users for home media use have totally open permissions

As a designer by trade, I guess I was looking more for an elegant tutorial of sorts that has well organized information. No offense towards cyberj0ck, as he seems to be a patron saint of freenas forums, but his tutorial looked like a 5th graders powerpoint. No table of contents? Come on. But, wouldn't it be nice to have something similar with an overview of possible configurations? Moreover, the only books on the subject seem quite out of date... I can't even buy my way out of disorganized materials.

The freenas docs are nice, but what they don't provide for someone like myself are real world examples based on general scenarios (home use, office use, off site backup, multi-user, single user, windows, mac, time machine, and so forth). Is it really too much to ask to have some kind of compendium of the collective knowledge of the forums? I know the forums are still useful, and I'll still refer to them, but it would be quite nice to not have to look at 17 posts to half-gather an idea of what I need to do. Moreover, many posts have multiple possible solutions posted to the same question. How do I know who knows what they're doing and who doesn't?

Also, it seems that I'm not the only one looking for this. It seems at the end of every blog post it seems there's someone out there asking for a pdf...

What I was looking for myself, and what became increasingly confusing, is determining how to set up a freenas that will both be used at home as well as an off-site backup from work. So, setting up something which is secure enough to be "out there in the open" as well as versatile enough to handle home media streaming needs. The discussions of users, groups and permissions is confusing, and I don't want to make mistakes which will haunt me down the line... after all, this is to make everything more seamless, secure and safe, no?

Why is there no PDF of this? Wouldn't it make sense and be a better use of everyone's time, especially the forum moderators who are probably exasperated and contemplate murder/suicide scenarios on the regular?

I'm offering to compile it as a pdf / e-book (or multiples thereof) with chapters and reference if someone is willing to provide the data.

I know there's no one size fits all to this, and it goes deep and there's a steep curve. I'm not saying I don't want to learn or invest energy and time into details and the finer aspects... I just think some organized help would be beneficial to all. I want to lay the groundwork for a fruitful experience with this hunk of steel that's made a home for itself under my desk.

If I'm wrong and misguided, lay it on me.

 

[depasseg]

Are you thinking like some sort of recipe book? I think I understand what you are looking for. For instance, the manual will explain what the options mean, but doesn't provide a step by step list of tasks to perform to configure a full capability.

 

[BigDave]

If I am reading you correctly, what you're asking for would take time. Time of course
is money. FreeNAS is open source software and the folks in this forum don't get paid
for the contributions they make. Your comment regarding cyberjocks effort being at
the level of 5th grade caused me to see RED.
Training is what you need. It's not free, but you get what you paid for right?
Or you can spend the next three years of your life in here, amassing as much knowledge
about FreeNAS and ZFS (as cyberjock and others have done) and write this super
organized how to guide for the benefit of people like yourself. I'm really sorry if my
comments offended you, but you insulted a man I have a lot of respect for.
Dave

 

[mjws00]

The problem with such a document is scope and keeping it current. A modest document to maintain may help very few for the tremendous amount of work necessary.

I like this idea in theory, but short of a nice fat cheque from O'Reilly who has time and knowledge to execute on this? A forum lets us share the workload and address the vast array of issues as they arise.

Something in the FreeNAS for dummies line could work. But BSD and ZFS geeks are likely too elitist to read it. ;)

Most who could write such a book carry pretty heavy workloads outside FreeNAS. So huge undertakings are a tough sell. But many will pop online quickly to help with a specific challenge.

There are a number of best practice guides that would benefit the community. They are exceedingly difficult to pull off well.

 

[danb35]

@BigDave, there's a big difference between insulting a person and saying that a particular piece of his work isn't very good. "Cyberjock is an idiot" would be insulting him. I don't think I agree with OP's assessment of the powerpoint (I don't see that a table of contents would be especially helpful, for example), but I also don't see it as insulting to cyberjock.

OP, I understand your frustration and have shared it. The official docs are very good, IMO, but they still focus more on "how each control works" than on "how to do X". I question, though, whether it's going to be possible or practical to create and maintain an "official" guide to "how to set up your FreeNAS". There are a few specific concerns I see with the idea:

• There's lots of different things people want to do with a FreeNAS box, so such a guide would necessarily be very large
• Software changes frequently, so "the right way" to do X today may not be the right way to do X in six months
• There's usually more than one way to do things, each with its own pros and cons. Take a look at the Zoneminder thread in this forum for just one example: there are people installing from source, people using binary packages, and people running virtual machines of completely different operating systems. There are often wrong ways to approach a given task, but multiple "right" ways depending on specific needs.
• Often a particular task depends on resources outside of FreeNAS. For example, using your home server as a replication target for your work server should ideally be done by establishing a VPN connection between your home router and your work router. Doing that should be well outside the scope of any FreeNAS docs

None of these concerns is necessarily a deal-breaker, but they would all make for a major project.

 

[BigDave]

@BigDave, there's a big difference between insulting a person and saying that a particular piece of his work isn't very good. "Cyberjock is an idiot" would be insulting him. I don't think I agree with OP's assessment of the powerpoint (I don't see that a table of contents would be especially helpful, for example), but I also don't see it as insulting to cyberjock.

Insult: to do or say something that shows a lack of respect for someone.
You're going to have to sell that somewhere else, you are just talking about degrees of insult.
The OP could have commented about the lack of a table of contents, no one would have a
problem with that, but comparing a grown man's volunteer efforts to that of an elementary
of education is an insult by definition. I call 'em, like I see 'em.

 

[cyberjock]

For the record, I didn't find any aspect of this discussions disrespectful to me.

I've thought about something like this. The problem this thread is talking about is because of multiple problems:

1. There are almost always multiple ways to accomplish a goal. Each provides advantages and disadvantages (for example, rsync versus zfs replication for backups).
2. Anyone can cookbook from a manual, but will you understand what you are doing?
3. Anyone can follow some steps, but are you capable of learning the material, then applying that knowledge?

To be brutally honest, few people here seem to be good at #2 and #3. So I can write a book about how to do a specific scenario, let's say a dataset for home, then a dataset for work. The work dataset is a backup from your work server using rsync. So if I don't share the work dataset but share the home dataset in my guide, people *will* ask "how do I share out the dataset for work". Then you'll also get the question "how do I put the work dataset as a subdataset for home?" and then I'll get the "what about if I want permissions for me, but not my wife, and another "what about permissions for my co-workers that visit my house, but not my wife and kids"?

Then I'll be asked how to do the backup with snapshots and replication, then I'll be asked how to keep the snapshots for only 2 weeks, then how do I rsync one directory, but do zfs snapshots for the rest.

See how fast things went totally out of control? Just the short example I gave would probably be about 100 pages, because you not only have to show how to do it, you have to teach it in a way that the user can understand and then apply.

Then people are going to ask why I did rsync instead of zfs replication, and next thing you know you have a 1000 page novel that talks about the history of the universe because people asked the questions.

There's 10 ways to do every thing you want in FreeNAS. There's pluses and minuses to every single choice. So it's nearly impossible to break down FreeNAS into something that can be taught to an 8th grader in a short book and expect it to work out. Unless you make it excessively thick, explain every advantage and disadvantage, and basically provide the equivalent education of an MCSE for FreeNAS, you're going to have only a small subset of users that will be able to use the guide. The rest are going to tell you what a dirtbag you are for not explaining other things in more depth, for failing to tell them how to handle whatever weird situation they can come up with, etc.

Sounds great in theory, it's a full-time job in practice. So unless/until people start saying they'd pay $50 for a book about FreeNAS that is 1000 pages (or more), it's not likely to happen.

I started writing a book, and I planned to go into as much depth as I could. I stopped at about 50 pages or so and it took me about 40 hours to do that much. That's not with any kind of proof-reading, fact checking, etc that needs to come later. I was told by one of the founders of iXsystems that if I wanted to go the book route I shouldn't expect a single copy to sell for at least 18 months from the day I start. T

Additionally, the big problem I felt with the whole idea of a book is that the problem is that this community is mostly sold on getting stuff for free. I have no doubt most users are going for the couch potato plugin, the sabnzbd plugin, the torrent plugin, etc. The number of threads that cover just those topics is mind boggling. I'd be stupid to expect that a community like ours would provide a return on investment on a book that I could easily spend a year or more writing. Most people would simply pirate it. As it is, most people around here take the OS because it's free, they try to put it on hardware that cost them nothing or almost nothing, and when they get stuck they aren't about to waste their time with forum searches. They're just gonna jump straight to posting a bunch of threads. If they are told to upgrade their hardware they flog those with 1000s of posts and try to tell us we are wrong, we're stupid, and we don't know what we are talking about. Nobody can promise any kind of return on investment on a book with that kind of crowd. It's disappointing, but it's the truth.

I've put dozens of hours into my newbie presentation. I've made no money off of it (I didn't want to make money on it to be honest). I also am embarrassed to say that it's not upgraded to 9.3 yet, but I'm thinking I'm going to rectify that today!

Those that know a lot about FreeNAS are probably working for the project, and therefore are plenty busy. Those that aren't involved in the project have this thing called "life" and "family" and probably aren't willing to devote a year or so of their life to it.

So unless someone is going to up-front the risk of the book if its a failure, expecting a resource of this kind of breadth and intensity, even if it wasn't free, is virtually impossible. Trying to get something like that for free is laughable, at best. I'd happily write a book if I was paid $10/hour to write it and I got a percentage of the profits. Unfortunately, that's asking *way* too much unless someone wants to crowd-source the book. Hmm... that's actually not a particularly bad idea.... Offer an electronic copy of the book and a paper copy for different prices. Hmmm....

The other downside to the whole book-writing thing is that, even if someone like me started a book right now, by the time it would be finished 10.x would be coming close to being out. 10.x is a major rewrite to FreeNAS. The WebGUI is nothing even remotely close to what you and I are used to, so that would require significant effort to rewrite the book. Would you want to throw $30-50 at a book that is likely to be release after the next version is out? The theory would likely be very very similar, but the actual WebGUI screenshots would be totally different.

Anyone on here ever do crowdsourcing before? If so, message me.

 

[Jailer]

Great post. Your post above providing an answer is more effort than the average person is willing to put into a question. :D

Actually a living document like wikipedia would be a good option for an undertaking such as this. Multiple users adding information and validated by the community. It would of course have to be administered for accuracy but that would take a lot less effort than trying to write the whole damn thing yourself.

 

[Ericloewe]

The other downside to the whole book-writing thing is that, even if someone like me started a book right now, by the time it would be finished 10.x would be coming close to being out.

This is where the pfSense book is. And they don't have a forum where one can just ask for the correct stickies and read them.

To be honest, the forums have plenty of high-quality material on the important topics not covered by the manual - hardware, networking basics, permissions (in the future) and ZFS concepts. It would make a lot more sense to clean the various stickies up, organize them and centrally list them than to attempt to write a monolithic book for what is a moving target.

The single point that I'd say requires more documentation is Permissions - Cyberjock's been working on a now mythical guide for quite some time now, but it is a gargantuan task, in good part due to Samba, which is an unholy mess that somehow mostly-reliably bridges several layers of complexity across OSes. The previous points still apply there - Samba is frequently a moving target that's inappropriate for a static book.

Great post. Your post above providing an answer is more effort than the average person is willing to put into a question. :D

Actually a living document like wikipedia would be a good option for an undertaking such as this. Multiple users adding information and validated by the community. It would of course have to be administered for accuracy but that would take a lot less effort than trying to write the whole damn thing yourself.

I'd say it's what we have here, in a way. It would be nice to be able to centralize things a bit, but the important stuff that can be easily covered has been covered. There comes a point where you just have to say "we have features X, Y and Z that work like this, do this, have these limitations and require that" - not everyone is going to use ZFS replication over Rsync, for whatever reason, for instance. Past a certain point, it's only reasonable to expect the reader to use his own mind to figure out what they need and how it will fit in their network. If spoon feeding/risk alienation is required, iXsystems sells TrueNAS.

 

[BigDave]

For the record, I didn't find any aspect of this discussions disrespectful to me.

One truth becomes glaringly obvious here, your hide is MUCH thicker than mine ;)
I'm gonna take my righteous indignation
and go mow the lawn or something...

 

[DJ9]

Just a thought, in case everyone forgot. iX offers that "Introduction to FreeNAS" class online for free. ;)
http://www.freenas.org/freenas-zfs-training/

 

[cyberjock]

Well, we had a Wikipedia document. The problem was that it was often ignored by the user that needed to read it (not unlike a book or other guide we might write). It used to be managed by the community too. Anyone (literally, anyone) could edit the FreeNAS manual by creating a wiki account. We've moved it in-house because support from the community was pretty poor (sometimes in error, but that wasn't the norm) and it just made more sense for a bunch of reasons to move it to an in-house solution (like it is now bundled with the OS itself).

In the big scheme of things its important to provide some basic knowledge of what a particular tech is (zfs replication for instance) and provide instructions for how to set it up. The guide provides basic knowledge (it doesn't always make sense on the first read, and you may have to do additional reading to ensure it makes sense for *you*), and the guide provides instructions for how to set up the various features, what goes in the fields, etc. So without adding a massive amount of complexity, it is hard to make the guide "better".

Please keep in mind that FreeNAS is free. TrueNAS is where the money comes in for FreeNAS to exist. TrueNAS uses customer service contracts. Soo you call iXsystems and someone on the other line answers and they resolve the issue. So in theory you don't really need a manual, or should have to refer to it only rarely. TrueNAS does come with a manual, and it's virtually identical to FreeNAS' manual with the exception that it has extra stuff because TrueNAS has extra features that FreeNAS doesn't have. Most customers will do initial setup with iXsystem's support, and they ask questions and see us do setup. So when they want to add a dataset, a user, etc they already saw us do it once. So paying for *support* really does pay off (pun not intended) with TrueNAS.

/shrug

 

[Luminousdolphin]

@depasseg: Yes, exactly. I don't want to configure without understanding. I was just looking for something better "curated" that relates to "best practice" as considered by forum moderators.

@BigDave: No insult. I was taking a humorous jab, and in no way discounting his contributions. Its very clear how much he has assisted everyone here, even from a cursory look at the forums and general FreeNAS discussions across the web.

I was also highlighting a difference between those more "mechanically" inclined versus those more "aesthetically" inclined, and their ability to organize their thoughts in a way thats better digestible by the masses. I am pretty certain this view is shared by many.

Also, I have seen a 5th graders powerpoint and its an accurate appraisal. I think cyberj0cks feelings are most likely more hurt by those asking for quick advice to do the wrong thing, the wrong way, and then disappear in a cloud of smoke from the forums without contribution. Moreover, I feel he wants the best for the FreeNAS community. My intention of the post was to encourage, and also offer my assistance, in creating something that I feel would be of maximum benefit to everyone involved, including cyberj0ck.

And yes, it is open source, and I'd venture a thought that the time spent crawling the forums and responding to the same questions over and over may be a less effective use of time and already meager resources than what I was proposing. It may save everyone a considerable amount of time to direct traffic to a resource like that for most of the stuff, and focus resources on more complex issues. It could also help FreeNAS reach a greater audience.. I have a feeling many take a stab and cave quickly because it feels daunting. And yes, it is daunting, but this could make it feel LESS daunting. What I am suggesting is not to make FreeNas "easier" as that's not possible, it's to make acclimation less life-consuming (its not uncommon to see people here spending ungodly sums of time trying to just get the box working!).

Moreover, did you catch that I was offering to assist, for free? I've laid out books for publication and I can assure you that doing so feels a lot easier to me personally than picking through the forums with a stick.

@mjws00 yes I agree. However, my thought was that the collective time spent replying to minor, specific instances compiled over a given period of time would be greater than assembling resources and cleverly attacking a broader solution.

@cyberj0ck thank you for your thoughtful reply.

My reply to your three qs is:

1. I agree, but I feel that multiple "options" or "paths" could still be integrated into a text.
2. relying solely on a cookbook will not guarantee any kind of deep understanding, but it will plant seeds and provide jumping off points. I am thinking more like the FreeNAS docs, but with added scenarios. Moreover, I think that many don't give themselves a chance because they are easily discouraged at first. Giving people a way to at least get things up and running and feeling accomplished could go a long way in the proliferation of FreeNAS.
3. I'm not saying just "steps" I'm talking about something more akin to a mathematical proof, where you can see the end result, how someone got there, the thought behind it, and retrace. For me, this is an excellent way to learn.

Your out of control example provides some points that I hadn't thought of in my initial question. I see the issue a little more clearly. It's kind of like the childrens books were you get to choose your own ending... however, to explain principles, followed by case study type examples could both help as a "quick-start" as well as contribute to a greater understanding of FreeNAS as a system.

I'd absolutely pay $50 for the 1k page book you mentioned.

I hear some swear by the 1983 "The Unix Programming Environment" book still to this day. Part theory, part reference and part example can guide people into thinking for themselves. This book had plenty of examples, and was also credited for being THE book on how to "think unix". (I believe the ubiquitous "Hello, world" example came from this book).

I'd also be happy to assist you in a crowdfunding effort. Basically all you need is a a number that you think you need, a bunch of goodies in separate tiers and a timeline. PS you could also crowdsource a free wiki or ebook! doesn't have to be a for sale paper book.

@ Eric yes also the stickies could use more organization. What I was envisioning initially was the "go-to" or "trusted" stickies in the how-to to get compiled into one PDF. And yes, permissions is where I'm particularly bamboozled. At least its a place to start. If i'm provided with categorized stickies with trusted info and screenshots I'd be happy to compile, as I said before.

@cyberj0ck TrueNAS was a thought of mine as well. But, as I said, I was eager to learn something new. Can one want to have a more involved guide with examples, and also want to learn and apply knowledge as well? I think that's reasonable enough. It should not be so black and white, no?

Curious to see how this thread evolves. Thank you all for your insight.

 

[depasseg]

My 2 cents: I think a book will be tough (hard to keep current, who wants a book, even an e-book?). But a Wiki or a paid "Premium supporters" section of the forum is something I would certainly pay for. I'd love an organized compendium of best practices. Right now I have a folder full of bookmarked forum posts, and I still haven't mastered my file permissions.

 

[SweetAndLow]

I keep a personal log of best practices every time learn something new or configure something new. I started off with one and now have several and they can be used across several versions because the idea are the same. I think a wiki or source control would be best because we could start off small and add more as more people join. Keeping it small will limit the amount of deprecated material and source control could help with versioning.

I also think creating this type of document is a crazy amount of work and needs formalities to make it successful.

 

[Luminousdolphin]

For example. My personal situation:

Hardware:
I have 16tb of wd reds inside a freenas mini with 32gb of ram, no ssds but have no issue adding if there are performance/stability gains. I set it up in raid z2 as one volume initially. I'm still using the router assigned IP and haven't set up ddns yet or any other functions. I have not set permissions or attempted to put any data on there, so I can start fresh if need be.

My needs:
I have a macbook I travel with to various locations, and a mac at home. I'm using the nas mostly for 2 things: remote backup from an offsite work machine, or when I'm working remotely via laptop (so everything is stored and current and duplicates aren't created on numerous removable drives which happened in the past) and hopefully serving media within my home and as backup for some important personal photos, documents and the like. Does this mean AFP is a no brainer? I know I will need one for an iTunes server, and one for a time machine server if I decide to use that. I see some use NFS. I may have windows machines attach. What should my shares look like?

Ideally I'd like the network drive mapped permanently, rather than FTPing in there, so that I can just drag and drop files through the finder.

What would give best throughput and what is the best compromise of results?

i.e. "What would cyberj0ck/forum mods do?" and list an option or two that APPROACHES best practice. I realize there is no true ideal.

possibility a:

versatility: 4/5
security: 3/5
speed: 5/5
caveat: problems with such and such
caveat: no support for such and such

and so forth...

It will stay at my home, but receive remote backup from one of my work locations.

What is a "permissions starter pack" for me to get moving with lets say 2 users, myself and someone with access to a given folder if I need to make a certain file or two accessible to them.

What is the most secure way to connect to my machine over the internet?

Speed remotely is probably a non-starter as its dependent on whatever connection i'm on at the time.

 

[cyberjock]

I know what *I* would do with one FreeNAS server. I also know what *I* would do with two FreeNAS servers. (Hint: They aren't the same.)

When it comes to the caveats, I don't think it's reasonable to be able to drop a list of them. It's really important that *you* understand what a given tech brings to the table, and when it performs slow it is important that *you* understand the tech enough to do some troubleshooting to figure out the problem, yourself. No matter what, you cannot get around the whole idea that the person *must* understand the tech they are using.

This isn't like a desktop. If your game is slow you can't say "damn, that's my 3 year old video card for ya". Instead you are left with, for example, slow CIFS throughput, and there's a dozen or more mechanism that could potentially be at fault. You *must*, without a doubt, understand enough about the tech to know when to apply it, and when not to. There is no getting around this requirement.

 

[Rilo Ravestein]

next thing you know you have a 1000 page novel that talks about the history of the universe

I would LOVE to see your view on that ;)

Maybe call in @jgreco too! :D

Might become a best-seller!

 

[BigDave]

@Luminousdolphin
My comment from yesterday was an over-reaction, my sincerest apology.
I will watch (with interest) your future contributions to our community
and from the stand point of a basic user of FreeNAS, will find some way
to make a personal contribution to your efforts.
Dave

 

[Luminousdolphin]

@cyberj0ck I think the main disconnect here is a difference of opinion that a thorough guide would encourage participation, knowledge, and a deeper understanding of the mechanisms at work rather than detract from it. As always, there's never a magic bullet. For myself I know that at least a guide that includes some detail of the thought processes involved and examples would assist in my full adoption and dedication to understanding freenas. Perhaps I am in the minority though.