I doubt it because ZFS has no mechanism to shrink a vdev.
No, this is for new disks / pools only. You can get bigger, but not smaller.
I havenât checked RC, but in beta at least, youâre aware that this documentation was incredibly incomplete right? As in missing everything except the newest methods.
Will this be resolved with 25.04 release, or only with 25.10 or later?
How does the upgrade path look like for people running VMs for production workloads?
Thank you for some more details.
I would never run production on a âRelease CandidateââŚ
Yes, but the point is to try⌠to see if the release, because in theory a release candidate is a candidate for release, could run production.
In TrueNAS land, RC seems to be beta (feature complete, needs testing) and beta seems to be alpha (early access, not feature complete).
So, I guess this means that there is no âproductionâ VM capability in the 25.04 train, and those who need VM support for production work loads will need to stay on 24.10 until a future release.
And Iâll preempt claims that the existing VM support was not âproductionâ either, it was not marked as experimental as above.
I upgraded a 24.10 server to RC1 mostly to see how the Onedrive cloud sync had been restored.
It took me about 15 seconds to recreate the credentials and set it up and it worked perfectly. It happened so fast I thought I had done something wrong. I seem to recall way back when while it was still working it took a lot to get the credentials in place. Well done. This was something I was looking forward to in Fangtooth.
Updated my test VM to 25.04-RC.1
- still no way to specify a MAC address for a NIC on a VM (afaik) NAS-134282
- PCIe pass through looks promising
- disk add dialog looks good
- still couldnât find an Ubuntu â24.04â image in the catalog
- concerned that there seems to be no way to run non NVME supporting OSes.
- I liked the CPU and memory fields on the VM init page. Much better than the options in Scale.
- What about CPU type? Is it just pass-through? What about a custom option, like there used to be? In the past I have had to downgrade the CPU to get a VM to boot⌠so that I could remove a faulty driver⌠etc.
Incus status is probably worth another thread. Weâl start one after we have some feedback on RC.1
In general, weâve moved away from âproductionâ as a term. Instead we use the software status page and its rankings from Developer, Tester, Early adopter, General, Conservative, Mission Critical.
âEXperimentalâ is our way of flagging that we donât think the feature will get to that Conservative or Mission Critical ranking, even if the general release is there. Its a major new piece of software that needs maturing.
However, it will improve with bug fixes. Weâll track its status.
For those of us on the Conservative rank, can I kindly ask for an upgrade path between 24.10 and 25.10? Although limited the VM feature in 24.10 is mature and Iâm not comfortable to migrate to a version where this feature still needs maturing.
I do not speak for iXsystems but if you think back to other major releases, areas of biggest change usually need time to mature.
For example, few folk here have discussed areas of pain re ZFS operation. But there was comparatively a lot of pain transitioning from 24.04 to 24.10 re: kubernetes â docker and I doubt it will be any different re: incus, VMs etc from 24.10 to 25.04.
Whether itâs incomplete documentation, quirks in the UI, whatever, youâre likely better off waiting for the 25.10u01 if big changes are being made in an area of TrueNAS youâre really relying on. It also really helps when other users have put together how-to guides for workarounds.
So if I had mission-critical VMs, etc. that are going to be affected, Iâd wait until the most common issues have been addressed or workarounds identified. Ditto scripts for emails, etc. ⌠thatâs usually u1 not the .0 release.
You cant delete root disk.
stgraber4:34 pm
The root disk must always exist and is the only one that gets snapshotted, backed up and moved around as part of the instance.
For VMs, you can make it tiny and never use it, instead relying on external disks (custom block volumes) for everything, but you canât fully remove the root disk as itâs assumed by a lot of different code paths to be present.
Admittedly a really minor thing, but in 24.10.2, when Apps are âUpdatedâ the button to compete this task requests user to âUpgradeâ. That is, âUpdate Allâ in the Apps section results in a pop-up confirmation UI requesting if the user wants to âUpgradeâ a given App. I know its minor, but Upgrade and Update arenât the same thing, has this changed in 25?
Given that you need to recreate VMs anyway, the easy upgrade path is:
24.10 â 25.04 â 25.10 , and only then recreate VMs and attach the zvols
There is a ticket in the backlog to address this, but it is not currently changed in RC.1.
Smooth upgrade, no fireworks.
Few notes, there might already be tickets.
The CPU temperature graph is per-thread? And is
Cpu
supposed to represent the CPU Package? Also seems like a lot of lost space.
The Apps Block I/O dashboard panel graph seems to be cumulative instead of representative of current / sampled values. Seems weird to use a graph for this if itâs cumulative.
I believe itâs been this way for a while and not really âimportantâ to me, but I canât see per-thread being useful. Not sure if thereâs an easy to way to only show core temps. The results of it do amuse me, though:
Again, amusement purposes only
I was wondering what the additional CPU was. I think youâre right that itâs the overall CPU temp.
I found it! Itâs listed as Ubuntu Noble.
Iâll start by asking is there a reason we donât have the full list at images.linuxcontainers.org available in the UI? Are these just the images that iXsystems has vetted at this point in time? Cause the options seem a tad random, we have OpenWRT, but no RHEL based option? Are there plans to expand this list with live updates, or will we be waiting for the next point release?