Hammy Havoc

Looking forward to giving this a try! Booting from internal storage is definitely interesting. Hope to see NetBird one day get first-party support like Tailscale does, that would be very cool.

What's the state of play with Unraid remote syslog in 2026? I've been trying to get it working with Alloy, but no luck, think it's down to how they're formatted.

Yes, it being exclusively available for SMB feels quite archaic. Likewise, with ZFS, I hope that we can just trigger a snapshot to be taken of certain shares/datasets one day and then avoid needing to use a traditional trash dir (that would be cool).

Has anyone gotten any closer to figuring this out since July? Element X is now unable to make calls because there's no LiveKit/Element Call, but can't set that up without MAS being set up. Would happily buy someone some beers if they can help me figure this out so my dad can talk to my seventy-something year-old aunt again who lives a thousand miles away in the middle of nowhere. :- )

Do you not feel that a premium file server distro should have a backup utility in the OOBE? No wrong answer here, just seeing what a mod thinks about the lack of implementation in the OOBE after twenty years. Personally, I consider this a necessary feature of utmost importance, and I'm always baffled when a release comes and goes but doesn't bring it. Sure, scripts, but the major selling point of Unraid is making things accessible to grasp via a great GUI, IMO. Not knocking anything or anyone here, just seeing if I'm alone in that expectation to be able to backup data when even every turn-key consumer NAS appliance has plenty of backup features in the OOBE.

They could kill two birds with one stone via a GUI for ZFS snapshot scheduling and zfs send in the OOBE, IMO.

Any info about this item in the changelog? Resolved some issues with existing 1MiB aligned partitions

Not yet, but very interested in any information anyone else has about it!

Sounds interesting! I run an LSI SAS2008, which as far as I know, should be something supported by this. I've been running Unraid without any issues with my LSI SAS2008 for a few years. What are some features/functions of storcli64 that I and other Unraid users may benefit from that we're otherwise not getting in the Unraid OOBE? And are there any gotchas? E.g., having the array stopped before doing anything using it? If so, this should definitely be a warning somewhere.

Has the "Advanced View" for the Docker page gone walkies in 7.2? I can't see it anywhere.

Flash backups should be a part of the standard update process itself by default, IMO. Reasonable defaults like that are something that I hope to see more of in Unraid. Likewise, up to 7.2 and 20 years of Unraid, but still no backup facility in the OOBE. Absolutely not a sexy or fun project, but perhaps one of the most important for a distro aimed at file storage. Not knocking anything here, but I'm always perplexed when a new version brings plenty of new features, but zero progress on a backup facility. Even if it was just a way of easily assigning internal, external and remote targets for ZFS replication then throwing in a cron schedule per backup task, it's absolutely better than how it currently is. At least then users could repair a ZFS dataset if they redundantly replicate it elsewhere.

What's the recommended way to implement MAS (Matrix Authentication Service) on Unraid if you're running this Synapse container? https://element-hq.github.io/matrix-authentication-service/index.html

Thank you very much! My ad blocking plugin didn't seem to like it, so I was oblivious to it. Disabled it for local IP addresses and can see it just fine. Excellent! :- ) Quality-of-life just increased hugely!

It's very laborious moving files around via File Manager, but the performance of using Krusader within a Docker container is sub-optimal versus via File Manager. Ideally I'd be able top open up two directories independently of each other and browse through both simultaneously, allowing me to transfer files from each easily, and hopefully also queue them: https://forums.unraid.net/topic/190851-allow-queuing-jobs-in-the-file-manager/ This would make transferring data to/from external disks or other servers much less painful.

It should absolutely be baked into the OOBE. If the scripts exist, it just needs some vars injecting from fields in the GUI if the scripts are as acceptable a solution as you feel they are. Anything that manages data as a solution should of course ship with a first-party-supported backup facility. Even if the backup is just a GUI for ZFS snapshotting, replication and automatic self-repair from redundant replicated datasets. The selling point of Unraid is in how user-friendly 95% of it is. The remaining 5% being less sexy yet still very important stuff like this.

Not the solution you're asking for, but using br0 as the network for the container gives you your own IP at the LAN level, similar in concept to a virtual NIC. Ergo, free ports galore.

I could well be wrong about this, but someone else mentioned recently that instant file moves only apply to individual shares on individual disks, regardless of filesystem and status as a dataset. It's why people recommend that instead of having a separate share for different kinds of data, you might consider a single share called data and reaping the benefits in perf for instant transfers within that share within each disk. Where that gives you issues is in ACLs. Unraid pushes you to create shares to manage access to data, but that model really opens up some problems if you have multiple users and everything in a single share, IMO.

Needs a proper secrets system like Komodo Core.

Herein lies the problem. You as an individual likely don't have the warchest to fight legal problems that may come your way. Storage is cheap—local or cloud. I'm all for decentralisation, but accepting anonymous file uploads is asking for problems. Where there's people, there's crime. If someone uploads CSAM to you and anonymously tips off the authorities with the decryption key then you're quite likely toast, bud. It sounds like a viable alternative to SWATing someone—land them in legal hot water through data. That and I'm not being funny, but even close friends and family—statistically speaking, the world is full of unpleasant things done by very average people, the chances of being related to one isn't impossible. I would think twice before even letting a friend use my Nextcloud instance for syncing photos from his phone versus encouraging him to set up his own server and instance so we can federate and maybe use zfs send, but even then, other than maybe immediate family members, I would be wholly reluctant, despite encryption.

Not the solution you're asking for, but starting the array in maintenance mode speeds things up significantly. But yes, I agree, I have the hotswap bays galore, but have never been able to hotswap using them.

Well, think of it this way—what is a plugin if not a nicely presented script(s)? Someone has to maintain the integrations with each platform, just like a standard script. I think what might make the most sense is if LimeTech provides a framework for community devs to hook into when developing integrations with other platforms so that it is a standardized thing—no need to reinvent the wheel for each platform. That then opens up the potential for the community to ask the likes of say Dropbox to maintain an official Unraid integration, and gives them control of their own integration repository, but also means that the community can pick up the slack when necessary.

Following on from https://forums.unraid.net/topic/190864-drop-checkbox-next-to-remove-on-plugins-and-instead-have-the-remove-button-present-the-user-with-a-modal-pop-under-confirmation-dialogue/, why not simply follow the UX of the Docker page where clicking an icon for a plugin yields an option to remove it? REMOVE is not going to be a commonly accessed feature. Likewise, the Support Thread could also be moved into that menu just like on Docker.

The checkboxes aren't a great design choice. The UI would be far less cluttered if we got an "Are you sure you want to remove %?" when clicking the button.

This seems to be what most "app stores" and program lists on various OS do. having it in a row in a separate column doesn't make a ton of sense.

This would elegantly pave the way for getting rid of this eyesore: https://forums.unraid.net/topic/190861-move-move-to-end-and-back-to-top-ui-icons-to-the-right-of-the-floating-footer-menu/