Improved Kubernetes stability for applications (NAS-125640).
Exposed Netdata UI under Reporting > Netdata for deeper real-time introspection and reporting on system performance.
The Netdata UI bases time on the local browser time, while the SCALE Reporting screen is based on the TrueNAS system time. Some difference in event time between the two screens is expected.
We can finally say that we can at-least support fresh installs of TrueNAS SCALE Dragonfish with our own OpenEBS backend.
We’re also working with HeavyBullets from Heavyscript for easy migration for existing users on a later date.
Checkout the DragonFish storage site on the website for more info if you want to!
Beware, there be, obviously, dragons and it’s still considered experimental till release.
Not sure if this is the place to record issues with Dragonfish, but it seems as good as any other place.
The old issue that plagued Cobia, of “App Services failing after boot if you are using a bridge” still persists in Dragonfish (see original thread in old forum here)
I’ve raised a ticket (NAS_128228).
Sad, as I was hoping this would be fixed with the new release.
Thanks for creating the ticket. I’ve assigned it out and we’ll see if we can get it fixed. We’re already past code-freeze for .0, but we’ll see if what we can do. In meantime, if you see these things, please by all means put the tickets in earlier rather than later. We’d rather duplicate tickets than something getting missed entirely.
Update: That ticket was a duplicate of this one. Scheduled for 24.04.1.
Yep, it is absolutely that. To date since RC1 went out, we’ve resolved roughly 137 issues found or reported against RC1. Handful of those still open are mainly Epic tickets for bigger items that don’t close out until SQA finishes their testing.
We tend to prioritize items that high impact or severity. I.E. things that hit 100% of people, or have severe consequences. Occasionally some lower priority ones slip in if they are super easy & low risk to merge as well. Doesn’t mean we get every single thing addressed for .0 though, some of them get punted into .1 if we don’t have time to resolve.
Thanks for the update.
I figured it may be a duplicate, but just wanted to make sure you knew its still not working in Dragonfish. Hopefully 24.04.1 finally fixes it.
This is more of a general issue that pam_session() will fail if user has non-existent home directory. If you had tried to SSH into server with that user or used any other service that tries to validate via pam_session it would have failed as well. TL;DR, if you want interactive sessions make sure your local users have valid home directories. You just noticed it was an issue for your server because you have configured homes shares and for historic reasons we require enabling samba to obey pam restrictions for this feature.
SMB users who do not have a “home” share are unaffected by this.
I’ve been using SCALE for a while and was on the 23.10 train and then yesterday switched to RC1 in 24.04 train and noticed a large increase in RAM with nothing else changing other than the version of course.
This is expected behaviour. 24.04 got better ARC management, before no more than half the available RAM was used for ARC (which left the other half kinda „wasted“). With 24.04 ARC can use almost all available RAM. See the 24.04 RC1 release notes
TrueNAS (far from exclusively!) has always had some trouble with clearing browser caches… Browser caches are a notorious example of one of the big three challenges in computing, caching. The other one is of course off by one errors.
We could thank those who wish to track everything we do on the 'net for various reasons (like adverts) for this incomplete cleaning of browser caches, and other nefarious longterm hidden data cache via those same browsers.
