TrueNAS 13.3 Wishlist

[morganL]

Any of the many scripts in the Resources area that handle backing up the config would be infinitely better for this purpose than having to deal with TrueCommand. But far better yet would be doing something useful with the backups you've already been making for the past ten years.

We're Ok with using other backup scripts. You can backup to your laptop or wherever.

If a user thinks their config is backed up on the system.. and then finds it isn't, that's a problem. Snapshots are not a backup. We only want to encourage safe processes.

TrueNAS does keep the config backups, but only for emergencies... and we don't recommend users rely on them since they can be wiped with a command or a hardware failure. This is the recommendation for CORE and SCALE.

 

[Redcoat]

since they can be wiped with a command or a hardware failure

Yes, this is what I have observed, but have not been able to determine the precise why's.

I have been a proponent of recovery of configs from the daily automatic backups, and described the process in detail for the distressed, but "recently" (the last year or so?) there have been multiple instances where the old config list has not been found - has seemed like a new situation but perhaps I just never internalized the frailty previously.

Is there no way that these early morning config backups can be "preserved"?

 

[morganL]

Yes, this is what I have observed, but have not been able to determine the precise why's.

I have been a proponent of recovery of configs from the daily automatic backups, and described the process in detail for the distressed, but "recently" (the last year or so?) there have been multiple instances where the old config list has not been found - has seemed like a new situation but perhaps I just never internalized the frailty previously.

Is there no way that these early morning config backups can be "preserved"?

The issue is that its not reliable... that's why we need to encourage another method.
Script, TrueCommand or manual.

 

[dak180]

Has this issue been experienced and reported? It should get its own ticket

My reading of NAS-108622 is that is exactly for this purpose; the link to the zfs on macOS people I think is just because they have documented the issue as a team that works with mac and zfs a lot.

 

[tiberiusQ]

• Enable Samba to process ElasticSearch API response NAS-118522

Makes it so search on SMB shares actually works (for macs).

Yes please !

 

[morganL]

My reading of NAS-108622 is that is exactly for this purpose; the link to the zfs on macOS people I think is just because they have documented the issue as a team that works with mac and zfs a lot.

Could be... we just haven't seen examples or other bug reports. Changing things (defaults) can cause compatibility issues and we prefer not to do unless there is pretty clear evidence we need to. If people have seen the problem firsthand when using SMB, then let us know.

 

[morganL]

There were 2 NVMe requests. Both are reasonable and we'd like to include in TrueNAS eventually. However, neither have yet been done in SCALE and we will experiment there first.

NVMe drive health is not as consistent has HDDs. Each NVMe drive type has different data protection and failure modes. Reporting of that internal health is spotty. The drives tend to either work or fail..... very unlikely to deliver errors. We'd be interested in reports of common failure modes that could be detected ahead of time. We've been shipping NVMe drives for 5 years and so far they have been very reliable.

NVMe namespaces create a new problem which is multiple "drives" failing simultaneously and requires system knowledge to be able to ensure redundancy is adequate. Its quite a large job to make it work well and reliably.

 

[dak180]

Changing things (defaults) can cause compatibility issues and we prefer not to do unless there is pretty clear evidence we need to.

I would not actually argue for changing the defaults; for me it would be sufficient if these things could be set at all. Right now it is imposable to set these things (in the gui) because they must be set at time of dataset creation (and cannot be changed later) which means you have to do it via the command line and I would prefer not to have to do that.

 

[morganL]

I would not actually argue for changing the defaults; for me it would be sufficient if these things could be set at all. Right now it is imposable to set these things (in the gui) because they must be set at time of dataset creation (and cannot be changed later) which means you have to do it via the command line and I would prefer not to have to do that.

If its in the GUI, then we have to test with option enabled and disabled.. every release. Most users already think there are too many options to configure.

Its worthwhile doing a GUI option if there a good % of users that see issues/benefits with each approach. Seeing real-world issues would be an indication of the problem. Otherwise, CLI is a reasonable fallback.

 

[rahulligma]

Update Rclone. https://ixsystems.atlassian.net/browse/NAS-127744

 

[FrankNAS]

Jail manager other than iocage. If it weren't for the 5 commits in 2023 I would say it was dead. Last release was 4.5yr ago.
EOL/remove plugin system.
If it were more mature I would say also add containerd.

 

[victort]

Jail manager other than iocage. If it weren't for the 5 commits in 2023 I would say it was dead. Last release was 4.5yr ago.
EOL/remove plugin system.
If it were more mature I would say also add containerd.

BastilleBSD!

But in all reality it’s not going to happen. TrueNAS CORE might be in its last days of seeing new features like this.

iocage is solid I would say. No need to make many commits on something that works. Anything you notice that can be improved?

 

[morganL]

We're working through the last issues (it can take days to investigate and decide on a specific feature).

The unchecking encryption bug fix requests will be fixed. It has a workaround, but a fix is possible.

 

[blanchet]

I wish that iX adds support for bhyve/NVMe in the WebUI.

According to Jim SALTER from Klara Systems, VM runs really faster with emulated NVMe controller than VirtIO.

For storage-heavy workloads, the benefit of bhyve’s emulated NVMe controller is difficult to overstate—it produced massive throughput improvements that even a long-time KVM fan simply cannot ignore.

FreeBSD vs. Linux – Virtualization Showdown with bhyve and KVM

Not too long ago, we walked you through setting up bhyve on FreeBSD 13.1. Today, we’re going to take a look specifically at how bhyve stacks up against the Linux Kernel Virtual Machine—but before we can do that, we need to talk about the best performing configurations under bhyve itself.

klarasystems.com

Migrating to TrueNAS Scale is not possible yet because I run a TrueNAS X20 that supports only TrueNAS Enterprise (FreeBSD)

TrueNAS System Revolutionize Your Home or Enterprise Storage

Choose the perfect enterprise storage solution, from all-flash to hybrid storage systems with maximum performance and scalability.

www.truenas.com

 

[Patrick M. Hausen]

Yes, please. It's one option in the UI and one extra "else if" or "case/switch" in the middleware.

 

[morganL]

I wish that iX adds support for bhyve/NVMe in the WebUI.

According to Jim SALTER from Klara Systems, VM runs really faster with emulated NVMe controller than VirtIO.

FreeBSD vs. Linux – Virtualization Showdown with bhyve and KVM

Not too long ago, we walked you through setting up bhyve on FreeBSD 13.1. Today, we’re going to take a look specifically at how bhyve stacks up against the Linux Kernel Virtual Machine—but before we can do that, we need to talk about the best performing configurations under bhyve itself.

klarasystems.com

Migrating to TrueNAS Scale is not possible yet because I run a TrueNAS X20 that supports only TrueNAS Enterprise (FreeBSD)

TrueNAS System Revolutionize Your Home or Enterprise Storage

Choose the perfect enterprise storage solution, from all-flash to hybrid storage systems with maximum performance and scalability.

www.truenas.com

The TrueNAS X20 can run The Enterprise version of SCALE. We just recommend it be a planned event.. preferably with Dragonfish and ARC improvements.

 

[morganL]

One of the last remaining requests wa ACME DNS. The proposal is we do this with the SCALE ACM DNS Shell Authenticator.. much like SCALE. This enables a broader range of authenticators and allows them to change. User or Community provided scripts can be written and adapted.

ACME DNS-Authenticators Screens

Provides information on the ACME DNS-Authenticators widget and settings.

www.truenas.com

 

[morganL]

The final 2 requests are for features not yet in SCALE Dragonfish. They are both desirable and we'll be looking to make progress in Electric Eel and beyond. However, there is no code to backport to 13.3.

 

[Patrick M. Hausen]

The samba feature could be implemented by just compiling in the support and letting the community find out and document how to connect Elastic running in a jail. Could probably be done with a SCALE app, too.

 

[morganL]

The samba feature could be implemented by just compiling in the support and letting the community find out and document how to connect Elastic running in a jail. Could probably be done with a SCALE app, too.

It will be done with SCALE first... but it requires extensive testing to ensure no unintended side-effects.