FreeNAS and TrueNAS Unification Blog & Discussion

Savell Martin

Contributor
Joined
Jun 10, 2013
Messages
164
Awesome that is what I wanted to know. So even if I go stable now it'll upgrade to OpenZFS when 12 comes out.
 

mike-pt

Cadet
Joined
Mar 13, 2020
Messages
1
Anyone tried this yet? I just tried to create a new dataset and encryption is not working, is there a place where I can provide a key first rather than auto generation?

Error: Traceback (most recent call last):
File "/usr/local/lib/python3.7/site-packages/middlewared/job.py", line 349, in run
await self.future
File "/usr/local/lib/python3.7/site-packages/middlewared/job.py", line 386, in __run_body
rv = await self.method(*([self] + args))
File "/usr/local/lib/python3.7/site-packages/middlewared/schema.py", line 973, in nf
return await f(*args, **kwargs)
File "/usr/local/lib/python3.7/site-packages/middlewared/plugins/pool.py", line 608, in do_create
raise verrors
middlewared.service_exception.ValidationErrors: [EINVAL] pool_create.encryption_options.key: Please provide a key or select generate_key to automatically generate a key when passphrase is not provided.
 
Joined
Jan 18, 2017
Messages
524
TrueNAS CORE 12.0 will go through the same ALPHA, BETA, RC1, RELEASE states that FreeNAS has gone through. There will be no changes to the software update process or the information available. The process will be more efficient within iXsystems, and we expect to deliver our users more with the same resources we have today. TrueNAS CORE 12.0 is planned for release in Q3 2020, which will be a much shorter development cycle for a release of this magnitude, thanks to this change.
Maybe I missed it but in regards to the TrueNAS Enterprise updates, they used to be a little behind the FreeNAS updates will this continue to be the case so we can find some of those edge case bugs before multi terabyte setup's run into them?
 

Kris Moore

SVP of Engineering
Administrator
Moderator
iXsystems
Joined
Nov 12, 2015
Messages
1,421
Anyone tried this yet? I just tried to create a new dataset and encryption is not working, is there a place where I can provide a key first rather than auto generation?

Error: Traceback (most recent call last):
File "/usr/local/lib/python3.7/site-packages/middlewared/job.py", line 349, in run
await self.future
File "/usr/local/lib/python3.7/site-packages/middlewared/job.py", line 386, in __run_body
rv = await self.method(*([self] + args))
File "/usr/local/lib/python3.7/site-packages/middlewared/schema.py", line 973, in nf
return await f(*args, **kwargs)
File "/usr/local/lib/python3.7/site-packages/middlewared/plugins/pool.py", line 608, in do_create
raise verrors
middlewared.service_exception.ValidationErrors: [EINVAL] pool_create.encryption_options.key: Please provide a key or select generate_key to automatically generate a key when passphrase is not provided.

This will be landing in the next week or so, sit tight. UI needed some work to fully enable encrypted datasets.
 

Kris Moore

SVP of Engineering
Administrator
Moderator
iXsystems
Joined
Nov 12, 2015
Messages
1,421
Maybe I missed it but in regards to the TrueNAS Enterprise updates, they used to be a little behind the FreeNAS updates will this continue to be the case so we can find some of those edge case bugs before multi terabyte setup's run into them?

Correct. We're still working out some of the specific details as to the lag time between Core / Enterprise, but we still are planning on doing this. Mostly we've gotta settle on how the upgrade trains will be handled, since they can be installed from the same base ISO.
 
Joined
Jan 18, 2017
Messages
524
Correct. We're still working out some of the specific details as to the lag time between Core / Enterprise, but we still are planning on doing this. Mostly we've gotta settle on how the upgrade trains will be handled, since they can be installed from the same base ISO.
good to hear, makes me feel better. I was thinking about it too and the only easy way was to have Enterprise report to a different update server than core. I have no idea how feasible that is however.
 

freqlabs

iXsystems
iXsystems
Joined
Jul 18, 2019
Messages
50
With the Alpha next week well now 2 days time, will it ship with OpenZFS 2.0? Or at least a version of it?
Also with current ZFS implementation once 12 comes out will it upgrade to OpenZFS 2.0 just fine?
I'm in the process of setting up FreeNAS so just wanna make sure I get it right.

The latest nightlies (and soon ALPHA1) are based on openzfs/zfs master, with a few additional local commits to add some features that have not been merged yet upstream, etc. It may be too early to call openzfs/zfs master "OpenZFS 2.0" yet, but the FreeBSD support is not one of the local commits anymore. :)

As for upgrading current pools, that will work but it is not required - remember upgrades are a one-way action.
 

Ericloewe

Server Wrangler
Moderator
Joined
Feb 15, 2014
Messages
20,155

MRBIQ

Dabbler
Joined
Apr 24, 2020
Messages
20
is there any plan to make trueNas enterprise OS so you came offer group of license depends on features of it , for example
E1 max ram 32G + max TB is 40 support 1 CPU
E2 Max ram 64 max TB is 80 support 1 CPU
E3 max ram 128 max TB 120TB support 2 CPU
And list of licence, I hope this will be available in future
 

Ericloewe

Server Wrangler
Moderator
Joined
Feb 15, 2014
Messages
20,155
Why on Earth would you actively want a byzantine system of multiple versions with arbitrary hardware limitations?
 

Patrick M. Hausen

Hall of Famer
Joined
Nov 25, 2013
Messages
7,667
I was about to ask the same. Arbitrary limits that have no technical foundation are just so ... Oracle? VMware?
 

Yorick

Wizard
Joined
Nov 4, 2018
Messages
1,912
Why on Earth would you actively want a byzantine system of multiple versions with arbitrary hardware limitations?

Maybe they're kinky that way? I try to not judge people's proclivities.

Kidding. I totally judge, of course.
 

HoneyBadger

actually does care
Administrator
Moderator
iXsystems
Joined
Feb 6, 2014
Messages
5,087
TrueCommand has a limit on the capacity/system count, doesn't it? So it wouldn't be unheard of to speculate on that.

But I think it was more along the lines of this question being asked from someone who's used to a by-capacity or by-featureset style license (eg: a backup product or legacy SAN) rather than the "one price gets everything" model that's more current. Or a hope that a lower-capacity system is a lower cost, and that someone wanting a small-scale system isn't looking at the same cost as a bigger buyer.

Or it could be that it's Oracle trying to worm its way back into OpenZFS. ;)
 

jenksdrummer

Patron
Joined
Jun 7, 2011
Messages
250
My thought would be something like this...

Paid-for support. OS is included. End of paid support w/ expiration, means additional features still function, but can no longer be configured other than to disable so it can move back to Core/CommunityEdition.

Paid support comes with 2 annual tickets. Cost is based on SLA. Engineer assigned and call back within 1hr is X (true priority enterprise-level support), then perhaps a 4hr (matches some vendors), 24hr (matches most non-critical), then 72hr (true non-critical).
 

Yorick

Wizard
Joined
Nov 4, 2018
Messages
1,912
Paid-for support

From the blog:
TrueNAS Enterprise will enable an extended feature set using a license key on supported platforms.

"Supported platforms" really is key. When you look at the requests for help on this forum, a lot of it, but a lot, comes down to enterprising, not enterprise, hardware builds. *badum-bish* From Realtek NICs to RAID cards to consumer boards with non-ECC memory to running FreeNAS as a VM on virtual drives (Lord Ha' Mercy!), if it can halfway run FreeNAS, people will attempt it. And that's great, that spirit of tinkering is definitely part of FreeNAS! It'd kill paid support though.

I am assuming that for starters, the "FreeNAS certified" systems sold on the ixsystems site will be a supported platform, and of course existing TrueNAS Enterprise gear. Beyond that, I wouldn't want to speculate. There could be a shortlist of gear that they'll support (these Supermicro boards with this memory and these Xeons; those HBAs; those 10G NICs; these SLOG devices), but that opens a can of worms even with a shortlist.
Obviously I am just thinking out loud, I have zero insight into what ixSystems is thinking.

Edit: For hobbyists, 90-plus percent of issues people encounter can likely be handled by a) ultra-orthodox hardware builds (Supermicro, Xeon/i3/Pentium, ECC RAM, LSI HBA, stop right there) and b) pre-planning pools (no, really, plan again) and c) climbing the learning curve with regards to "what is networking" and "how do Windows permissions work, what is authentication in Win10, how does that interact with Samba"
 
Last edited:
Joined
Jul 2, 2019
Messages
648
Better yet, license the OS to a single NIC MAC (!) (true story)
I can remember when Digital (DEC) used to license each of our Alpha workstations specific to the MAC. That was a loooooong time ago!
 

jenksdrummer

Patron
Joined
Jun 7, 2011
Messages
250
From the blog:


It'd kill paid support though.

Paid per-incident support would basically make it worth doing outside of just operating system sales. Microsoft, for example, makes their bread on support tickets, not operating system sales. The price point for the OS is such as to get a token 'buy in' cost more or less to size their support.

To that same point, there are options when it comes to a triage approach for paid-for-support; as an MSP the company I work for handles that rather gracefully. 3rd Party offshore for answering typical customer calls, that *quickly* escalate once the basics are covered.

Just tossing a number out there, lets say the model is $500 annually, with the CPU SN for socket 0 being the registration path. For that you get the enterprise features and 2 "free" support tickets. Additional support tickets cost $250. Not unreasonable in my book. I'm not sure of a safe way to disable the Enterprise features and not absolutely break something, but, if that can be done safely (for the most part) should one opt to not renew, then that makes a ton of sense. The renewals would be $500 in this example; and you still get 2 tickets annually. MOST of us probably won't need one, let alone both, but it shows a value there for those who license the product. Also, that's far enough out that budget-build-joe won't pay for that premium; not when likely Core Edition satisfies the budget ($0) and community-based support ($0, though potentially not as awesome (no offense, but we the community likely didn't contribute code))
 
Top