Working fine for the most part, except for the webVNC bug that won’t be fixed.
To ask have you confirmed that the cursor fix (275760 – bhyve: xhci not working on linux guests) is not in BETA2?
Like @victor says, still no mouse control in graphical VMs, so impossible to install any (Linux) distro that does not have a serial console enable, which is most.
Correct me if I’m wrong but I think the mouse movement works, you’d just need to refresh the screen to see if it actually did move.
I suppose that’s a more accurate description. I call it “no apparent mouse control”. Is this really how TrueNAS CORE will end?
Ahhhhhh
I saw that but that was mentioned specifically with Web Interface. I assume both an actual VNC client and the Web Interface still suffer the same problem but was looking for confirmation.
The underpinnings of VMs and Bhyve are inherited from FreeBSD.
If the version of FreeBSD has a bug… it will be inherited by TrueNAS. That is always the case.
So best to find which version of FreeBSD 13.3 has the fix… then we can look at using that in future updates. Do you know the version?
Just started testing 13.3-BETA2 and so far so good but I was a bit gutted to see the shell removal from the UI. I know it had its issues but for the odd command I found it very convenient. Any chance with enough support iX would bring it back?
@kris is unlikely to change his mind. The shell in CORE has been half-broken for years and @kris doesn’t have the resources to fix that issue (along with a bunch of other ones that don’t affect core functionality).
For example, we were told that the number of SSL certificate issuer options would increase beyond Route 53 in the 13.3 GUI. That hasn’t happened either.
AFAWK, 13.3 is basically frozen. Unless iXsystems pivots back to FreeBSD from Linux I doubt we will see much more than bug fixes between now and EOL.
Not even that. It also depends on the bug.
The Snapshots page is buggy, and makes you nervous to prune / manage snapshots, but it’s unlikely to be fixed, since it requires fixing the underlying GUI.
This is true for other sections of the GUI as well.
Where in dahell is the emoji wildly welding a hammer towards baby bunnies? CORE’s fate is a horrible situation, and most are drunk on the Kool Aid at this point so I highly doubt there would ever be a pivot. iX is burning bridges even if they did; I, personally, am implementing/testing my next setup (and I would suspect others are as well). If iX ever did pivot back why would anyone switch back to a known flight risk?
Companies will use this opportunity to evaluate what does and doesn’t work for them with TrueNAS and come up with a plan of what should be implemented next; if they already have to retrain on a UI, they may as well use the opportunity to implement a “better for them solution” and then retrain.
Allow me to quibble a bit.
@kris alluded to the efforts that iXsystems had to put into making FreeBSD work with changing hardware and software alike. IIRC, he mentioned this took away from their ability to deliver on software upgrades that the consumer could see because they had so much work to make the plumbing operate as expected.
I can understand why the engineering team prefers being able to focus more on the middleware and less on the plumbing. The current mindshare re: OS’ is firmly in Linux’ favor, despite all the security, reinventing the wheel, etc. issues that entails.
As a user, I simply want something that works reliably. TrueNAS usually delivers on both Linux and FreeBSD. If only the snapshot replication resume feature actually worked as advertised. Oh well.
Okay, so iX being an “appliance” don’t they better control the hardware? I’m sorry, I see that as a BS stance (I think their decision is based on money; customers want “containers” and iX wants the name recognition of “docker” and “Kubernetes”).
Getting the plumbing to work as expected is part of the problem I see; with systemd
changing the rules every other release you’d have your hands full just chasing rabbits (I call this a “box of rabbits” situation, everything is fine until you open the box). Kernel devs have been pulling their teeth out dealing with all the crap systemd tosses down the pipe for years. Not to mention every part of the Linux OS has dependency issues (every piece has a list of dependencies and that in itself is a testing/stability/security nightmare).
Um, no? Are you running TrueNAS on iX hardware? Because I’m not; I’m running it on my own, as are, I suspect, most of the folks here.
But it’s really tiresome to rehash the “iX should keep developing CORE” argument every month or two, so I’ll refer you back to the recent discussion on the old forum:
I am. FreeNAS Mini. I used BSD and wanted a BSD solution when my wife was pregnant with our first child (and I knew I wasn’t going to have time to play with servers anymore). I found FreeNAS. I purchased a Mini and paid extra to have them set it up. I thought this was my way of making a donation to them (I never got my Sickers or Beasty horns though!!). I did the same for my firewall too.
There are two forums?! I’ll have a look. Thaks.
12 pages?! *scoff* I’ll get back to you in a week.
In that thread I already see the container discussion. BSD jails are not containers (Example: “Dan, please send me your ngnix jail.”).
Below would be an example of a tool to look into if you wanted containers on FreeBSD. It has the ability to create simple jails like iocage but also to create containers and images (something you can share).
GitHub - bsdpot/pot: pot: another container framework for FreeBSD, based on jails, ZFS and pf
Still reading that ‘old thread’. All this talk about saving the GUI. Rip the head (GUI) off TrueNAS CORE as far as I’m concerned. I want stable and ease not pretty. If I had some scripts on my laptop that could run by double-clicking, cron, or what-have-you. What do I need a stupid graphic green or grey button to show me the status? Saves me one less login, tree to expand, or thing in my way to get something done.