TrueNAS SCALE 23.10-RC.1 has been released!

[eturgeon]

We are pleased to release TrueNAS SCALE 23.10-RC.1! This is codenamed "Cobia".

Early releases are intended for testing and early feedback purposes only. Do not use early-release software for critical tasks.

Notable changes:

• The legacy pool creation screens are removed and the new pool creation wizard is the primary UI screen for creating new storage pools NAS-123697
• Apps > Settings adds the Manager Container Images option for downloading, updating, or deleting specific container images from TrueNAS
• Web interface feedback buttons are visible for community feedback on early releases.
• As part of the netdata implementation and overhaul of the reporting features, Graphite support is no longer built-in with TrueNAS SCALE 23.10 (Cobia) NAS-123862.
• Fix for the UPS service to allow device detection NAS-123625

See the Release Notes for more details.

Release Notes: https://www.truenas.com/docs/scale/23.10/gettingstarted/scalereleasenotes
Download: https://www.truenas.com/download-truenas-scale
Documentation: https://www.truenas.com/docs/scale/23.10/

For any discussions in the Forum on Cobia-specific issues, please use the Cobia Forum: https://www.truenas.com/community/forums/Cobia/

Thanks for the early testing of TrueNAS SCALE Cobia! As always, we appreciate your feedback!

Also, if you find a bug, please create a ticket at https://ixsystems.atlassian.net/jira/software/c/projects/NAS/issues.

 

[Davvo]

ZFS dRAID Pool Layouts, amongst the other great improvements, are something that makes me a bit engious as a CORE user.

Is there any future plans to bring it into CORE, maybe after a "testing" period on SCALE?

Thank you for your hard work!

 

[LarsR]

Just did a second restart of my box and saw that i still had to run upsdrvctl start for my ups to work...

 

[morganL]

ZFS dRAID Pool Layouts, amongst the other great improvements, are something that makes me a bit engious as a CORE user.

Is there any future plans to bring it into CORE, maybe after a "testing" period on SCALE?

Thank you for your hard work!

dRAID requires pretty extensive testing of both the zfs software and the webUI.

TrueNAS CORE 13.1 will have the OpenZFS 2.2 which includes dRAID, but it won't have the UI. I doubt the UI elements will get into CORE during 2024.

We'd recommend testing it with SCALE and then letting us know why you need it in CORE.

 

[Kris Moore]

Just did a second restart of my box and saw that i still had to run upsdrvctl start for my ups to work...

That sounds like a ticket needed. We'll take a look and see if it can be addressed for .0 next month.

 

[Davvo]

dRAID requires pretty extensive testing of both the zfs software and the webUI.

TrueNAS CORE 13.1 will have the OpenZFS 2.2 which includes dRAID, but it won't have the UI. I doubt the UI elements will get into CORE during 2024.

We'd recommend testing it with SCALE and then letting us know why you need it in CORE.

Thank you for the information! As I see it, dRAID is a great option for ZFS users indipendently from the TN version... it's like asking (example) why we need RAIDZ3 on CORE when it's available on SCALE. If I am to remain using CORE since I prefer jails over kubernets, it would be nice to have dRAID. Bascially I don't see how the needs of it for CORE users would be different from SCALE users, tell me if I misunderstood you.

But I am not an enterprise user, and their opinion is (legitimately) what it matters... so I just hope than in a few years there will still be enough enterprise users to justify iX's effort in bringing this feature to CORE.

 

[Arwen]

The only real feature for dRAID is integrated hot spares. For home & small office users, using hot spares is probably rare. Not saying hot spares are not used by home or small office users. But, in my opinion, it's a lesser used feature for the smaller NASes, (and free users).

With the potential of RAID-Zx expansion, their may be less reason to use dRAID. You have to specify the dRAID vDev at creation, which includes how many integrated hot spares, what parity level and how many data disks. Since dRAID hot spares can't be shared between multiple dRAID vDevs, nor other pools, they are less useful in some cases. (Again, in my opinion...)

Though who knows what the future will hold?
Perhaps dRAID will get column expansion options.


PS: There was a discussion a while back asking what ZFS commands were to be used for dRAID column expansion. Turns out that forum poster was reading another vendors implementation of dRAID, which had nothing to do with ZFS.

 

[morganL]

Thank you for the information! As I see it, dRAID is a great option for ZFS users indipendently from the TN version... it's like asking (example) why we need RAIDZ3 on CORE when it's available on SCALE. If I am to remain using CORE since I prefer jails over kubernets, it would be nice to have dRAID. Bascially I don't see how the needs of it for CORE users would be different from SCALE users, tell me if I misunderstood you.

But I am not an enterprise user, and their opinion is (legitimately) what it matters... so I just hope than in a few years there will still be enough enterprise users to justify iX's effort in bringing this feature to CORE.

Understand the problem. However, TrueNAS CORE is now more focussed on quality/stability while SCALE is focussed on innovation.

Many current TrueNAS users prefer Jails (so do I - it is quality/stability), but the industry is moving to Containers (innovation). Developing the same software twice, before the tech is proven, is not economical. After it is proven and there is enterprise demand, we can consider, but Jails are rarely used in the Enterprise.

IX did the work to make sure all ZFSonLinux features also went into FreeBSD.. so the underlying ZFS remains the same. TrueNAS CORE 13.1 will have ZFS 2.2 which includes dRAID. There will nothing stopping the use of this via CLI for those that want to try it out before it is proven. I would suggest that dRAID only be used on larger systems (24+ drives) with probably a good amount of RAM. Jails in a VM on SCALE is also a choice. TrueNAS users will have many choices, but not every choice.

 

[Arwen]

...
TrueNAS users will have many choices, but not every choice.

Their is no one NAS to rule them all.

 

[LarsR]

That sounds like a ticket needed. We'll take a look and see if it can be addressed for .0 next month.

I tried to replicate it this morning with three separate restarts, but had no luck.

 

[Etorix]

Understand the problem. However, TrueNAS CORE is now more focussed on quality/stability while SCALE is focussed on innovation.

Thanks for your explanations.

As dRAID is only suitable to large arrays and is even less flexible than regular raidz (no column expansion, would require a further 24+ vdev to enlarge the pool), it is likely that very few home users have a case to merely consider dRAID at all, much less actually use it.
This is about CORE users not feeling "left behind" or pushed to migrate to SCALE—or to XigmaNAS.

Does Cobia features improved ARC management? This would be a much more relavant feature than dRAID for most potential users.

 

[Davvo]

@morganL understandable, thank you.
@Etorix I haven't seen anything about it in the release notes.

 

[Patrick M. Hausen]

Understand the problem. However, TrueNAS CORE is now more focussed on quality/stability while SCALE is focussed on innovation.

Understood and perfectly ok from my point of view. Still please consider syncing e.g. bhyve with FreeBSD upstream. There have been so many improvements and bug fixes ...

 

[morganL]

Understood and perfectly ok from my point of view. Still please consider syncing e.g. bhyve with FreeBSD upstream. There have been so many improvements and bug fixes ...

Suggest making that a separate thread around TrueNAS 13.1..... volunteers to thoroughly test would be appreciated.

 

[morganL]

Does Cobia features improved ARC management? This would be a much more relavant feature than dRAID for most potential users.

No major improvement in ARC for Cobia. There is ongoing work that we hope makes it into Dragonfish.
Cobia will be better for the SMB on NVMe storage use-case... where ARC is less important.

 

[Kris Moore]

Understood and perfectly ok from my point of view. Still please consider syncing e.g. bhyve with FreeBSD upstream. There have been so many improvements and bug fixes ...

Bhyve backend updates will be ingested as part of the update to later 13-STABLE FreeBSD base. However no MW/UI changes planned beyond that. Anybody seriously using bhyve does have my sympathies though, oh the stories I could tell.... Stop your suffering and use KVM already :)

 

[Patrick M. Hausen]

May I suggest you join Michael Dexter's bhyve production users call on Zoom to get an impression of the real state of affairs. Bhyve is perfectly production capable. I run our domain controllers on CORE.

 

[victort]

May I suggest you join Michael Dexter's bhyve production users call on Zoom to get an impression of the real state of affairs. Bhyve is perfectly production capable. I run our domain controllers on CORE.

I am in full agreement with this. I use CORE for our VMs also.

 

[Kris Moore]

We ran our entire infrastructure on Bhyve several years back. Dozens of physical systems and 100+ VM's. It had a lot of issues, especially when running non FreeBSD guests, never-mind our issues with FreeBSD on the host side across a wide variety of hardware. Granted its probably matured a bit now, but moving off saved us a ton of headaches, no desire to revisit that. All depends on your use-case though on what is suitable for your needs. This is speaking as a long time FreeBSD guy myself who didn't want to have to learn $OTHER :) But this is probably best for another thread/topic.

 

[morganL]

May I suggest you join Michael Dexter's bhyve production users call on Zoom to get an impression of the real state of affairs. Bhyve is perfectly production capable. I run our domain controllers on CORE.

I assume you are OK with Kris's response about "update to later 13-STABLE FreeBSD base"... start a new thfread if there are any specifics you need to confirm. Thanks.