TimeMachine not working no matter what I do

[dabrahams]

"The selected network backup disk does not support the required capabilities - please choose a differnt network backup disk"

I am stuck here with TrueNAS 12.0-U7. Ideas anybody?

 

[dabrahams]

This part, at least, appears to be a permissions problem of some kind. TM is able to select the disk and create a new sparse bundle if I enable "Guest Access" in Sharing > SMB > Edit > Advanced Options. If I then stop that backup, disable guest access, and unmount the disk, I can initiate a new backup, and data gets copied! This is awesome, except I really don't want to have to turn on guest access to begin backing up every machine. So what is it about guest access and bundle creation?

 

[dabrahams]

Oh, this is too weird. With guest access off, I rebooted the NAS and went to a different mac, selected the time machine share, and started backing up with no problems. #AFLOC

 

[Glorious1]

I don't have a fix (I'm stuck with Time machine too), but I wonder what OS version the two macs are running. Is the successful one an old version?

 

[dabrahams]

I'm succeeding with Monterey 12.1 on several macs. @Glorious1 where exactly are you stuck?

 

[Glorious1]

It's all laid out in the thread I linked to. My problem is not the same as yours was. I'm also on Monterey 12.1.

 

[dabrahams]

It's all laid out in the thread I linked to. My problem is not the same as yours was.

I realize that, but I get the impression that your problem evolved over the course of that thread, and I was hoping for a concise description of the current status, or a pointer to it.

 

[drjustice]

At his request, I sent a Debug file by PM to @anodos in October '21. He has not even acknowledged receiving it, much less any progress on this issue. To say I'm underwhelmed would be an understatement.

edit: I just saw that 2202.RC2 was just out, and checking the release notes, I see this bug has been fixed, but fixed how and why, I have no clue:

https://jira.ixsystems.com/browse/NAS-113246

I may try to update to 22.02-RC2 in a bit to test.

 

[anodos]

I work on issues people raise on forums as I have time (and my inbox can get quite full). mDNS issues can be somewhat difficult to reproduce in the lab (can rely on particulars of networking hardware in the environment).

I currently have about 5 different versions of MacOS doing continuous backups with time machine to Core and SCALE and mDNS is stable, and so I don't think this is an issue in general with avahi.

If you have multiple NAS interfaces on the same subnet, try hard-coding one interface name in the avahi config via `allow-interfaces=<list of nics>` and see if the issue goes away.

 

[anodos]

120 seconds is the TTL for mDNS service announcement. Your home networking should be acting as an mDNS querier, and keep this stuff alive, and so if your NAS is disappearing after 120 seconds you may want to take a close look at igmp settings. IIRC some vendors tend to break mDNS in exciting ways.

 

[drjustice]

I work on issues people raise on forums as I have time (and my inbox can get quite full). mDNS issues can be somewhat difficult to reproduce in the lab (can rely on particulars of networking hardware in the environment).

I currently have about 5 different versions of MacOS doing continuous backups with time machine to Core and SCALE and mDNS is stable, and so I don't think this is an issue in general with avahi.

If you have multiple NAS interfaces on the same subnet, try hard-coding one interface name in the avahi config via `allow-interfaces=<list of nics>` and see if the issue goes away.

I wasn't knocking the fact that it's not fixed -- simply the fact that you asked for a debug file, which I did send (long with more other information), but then I have not heard even a peep even when I reached out by PM or this thread -- not even acknowledgement that you got the debug file (nevermind a jira ticket which we could follow and add information to). As far as I knew, this issue had gone to the big old issue graveyard.

Glad to know the status, even if the current status is that it's in limbo because not repro.

 

[nathan47]

I was just running into this, and the post from @awasb here solved it for me.

Multi-user Time Machine purpose

Search finds a nice thread on this topic but it references something that is gone. So... To start... I'm new to TrueNAS. I created a Data set with nothing funny. Called it time_machine. It is case sensitive. I created an SMP share and with Purpose I put Multi-user Time Machine. It then...

www.truenas.com

Notes:
- Do not create an SMB data set
- When you create the ACL, in my case it was POSIX_RESTRICTIVE
- Step 6 appears to no longer be necessary (at least in TrueNAS SCALE)
- After step 3, I didn't immediately restart the SMB service, I instead waiting to restart SMB until after step 5