Unraid OS Version 7.0.0-beta.1 available

Since the Unraid array is optional, is it still going to be required to stop all the array/pools to do array/pool maintenance?  

 

I would like a true hybrid, where I can have independent arrays/pools, and mange them independently...which means I don't have to stop the unaffected arrays/pools.

 

I have one pool for media.

I have another pool for JBOD.

I have another pool for maximum data protection.

Anyone having issues with fstrim on Unraid 7?  Got an email with my scheduled trim:

 

cron for user root /sbin/fstrim -a -v | logger &> /dev/null

fstrim: /mnt/disk1: the discard operation is not supported

fstrim: /boot: the discard operation is not supported

 

When I run it manually it says this:

 

via cli:

15 minutes ago, meunraid said:

Anyone having issues with fstrim on Unraid 7?  Got an email with my scheduled trim:

Go to Settings - Scheduler - TRIM and make a dummy change then hit apply, that will update to the new script, and you should not longer see those.

 

The new script exists since v6.12 BTW, but with v6.12 fstrim did not try to TRIM the disks and flash drive, so it didn't cause the errors.

17 minutes ago, JorgeB said:

Go to Settings - Scheduler - TRIM and make a dummy change then hit apply, that will update to the new script, and you should not longer see those.

 

The new script exists since v6.12 BTW, but with v6.12 fstrim did not try to TRIM the disks and flash drive, so it didn't cause the errors.

Thank you

5 hours ago, Jaybau said:

Since the Unraid array is optional, is it still going to be required to stop all the array/pools to do array/pool maintenance? 

Although this would be nice to have I suspect it will not be soon as starting/stopping of services is tied to bringing drives online/offline.    Working out exactly which services are affected by a specific pool is likely to be a very difficult thing to do in a robust way.  

Updated to 7 beta and non of my external drives are recognized correctly anymore. Seems to be something with reading the serial #'s of the drive and duplicate (same model drives) only 1 would show... Does anyone have any ideas on how to fix? Works perfect when on 6.12.10.

 

Did a rollback and all drives are recognized and array is working again... Curious, will the final 7 version have this fixed?

 

Edited July 7 by dejected-fiddle3213

11 minutes ago, dejected-fiddle3213 said:

Updated to 7 beta and non of my external drives are recognized correctly anymore. Seems to be something with reading the serial #'s of the drive and duplicate (same model drives) only 1 would show... Does anyone have any ideas on how to fix? Works perfect when on 6.12.10.

 

Did a rollback and all drives are recognized and array is working again... Curious, will the final 7 version have this fixed?

 

Did you collect diagnostics when on 7?

Unfortunately, I did not. I can update back to 7 to take diagnostics if needed though. PM me.

Its impossible to change the Docker custom network type, it always saves as ipvlan if you change and save anything on the docker setting page. It always adds the dockernetworktype 1 to the cfg.

Has work been done on fuse? Just noticed i'm down to about 300Mb/s with user share to user share transfers.

Had about double that with 6.12.10.

1 hour ago, Yock said:

Has work been done on fuse?

AFAIK nothing has been done, though just the kernel jump sometimes has some impact, it should hopefully see a nice performance boost once LT can update to kernel 6.9 and use FUSE passthrough, did a test a couple of days ago and it's performing about the same for me:

 

 

 

1 hour ago, Mainfrezzer said:

Its impossible to change the Docker custom network type, it always saves as ipvlan if you change and save anything on the docker setting page. It always adds the dockernetworktype 1 to the cfg.

Thanks for the report, I can reproduce, though note that for now you can still use macvlan is you disable bridging.

2 minutes ago, JorgeB said:

Thanks for the report, I can reproduce, though note that for now you can still use macvlan is you disable bridging.

i do know. im testing it on purpose to see if the kernel panics still exist. Ive swaped to ipvlan and lxc container, for anything that needs its own ip, a while back because the macvtap has an issue with inbound package drops.

9 minutes ago, JorgeB said:

AFAIK nothing has been done, though just the kernel jump sometimes has some impact, it should hopefully see a nice performance boost once LT can update to kernel 6.9 and use FUSE passthrough, did a test a couple of days ago and it's performing about the same for me:

 

 

 

 

My transfer speeds really took a nose dive with Unraid 7 for some reason. Exclusive share speeds are still quick as ever.

Looked up the fuse passthrough yesterday, looks interesting and looking forward to it.

Anyone having problems adding/remove tiles from the Dashboard using Tile Management? It's just blank for me:

 

 

If I clear my cookies I get the default Dashboard back and all the tiles which I can re-order/remove, but can never then add them back later.

29 minutes ago, erf89 said:

Anyone having problems adding/remove tiles from the Dashboard using Tile Management?

It's a known issue, fixed for beta.2

Unassigned device issue, some files appear with unsupported reparse tag 0x80000017

Filesystem is NTFS compressed with lzx compression within windows 11 (command: compact /c /s /a /i /exe:lzx "*")

Mounting disk using Unassigned device plugin.

Problematic files marked in red in mc.

Please help.

 

n5105-diagnostics-20240709-0844.zip

24 minutes ago, sharicov said:

Unassigned device issue

then :

 

 

My Unraid server just seems to ... 'stop'. Drives stop spinning, can't connect too it, its like its not even... a thing anymore. Have to do a hard reset and I cringe every time. I'd love to upload my diagnostics for folks, I'll just have to figure out where to get them at.

zima-diagnostics-20240709-0032.zip

Edited July 9 by stry8993
Found and added diagnostics.zip

34 minutes ago, stry8993 said:

I'd love to upload my diagnostics for folks, I'll just have to figure out where to get them at.

Enable the syslog server and post that after a crash in a new thread in this same subforum, also upload complete fresh diagnostics.

16 minutes ago, JorgeB said:

Enable the syslog server and post that after a crash in a new thread in this same subforum, also upload complete fresh diagnostics.

Should I enable that now and just let it run until I crash again?

4 minutes ago, stry8993 said:

Should I enable that now and just let it run until I crash again?

Yes.

I just wanted to let you know that I have observed the following. Before I switched to 7.0.0 my nfs performance with 6.12.10 was between 850 and 950 MiBs with 7.0.0 only between 400 and 500 after a downgrade to 6.12.10 again at 850 and 950 MiBs.

On 7/1/2024 at 3:28 AM, JorgeB said:

If a share has files in two pools it can no longer be exclusive, you need to assess it over /mnt/user, or would not see the files from both pools,

 

I just noted what looks like a limitation of re-assigning shares and hope you can give some info here to help me determine if I should bring this up as a bug or feature request.

 

I've created a RAIDZ2 pool where I can make shares that get Exclusive access by default - Exclusive set in general preferences and the files are only accessed in the one pool, no secondary storage.

The limitation I'm seeing is when reassigning the location/pool for an existing share that is currently on the traditional array. Such a share has Exclusive = NO. That doesn't change to Yes when I reassign its location to the ZFS pool and set secondary storage to none. By definition it has the same settings as a new share (one pool only).

Can this limitation be remedied such that we gain exclusive access when making this change?

22 minutes ago, Espressomatic said:

 

I just noted what looks like a limitation of re-assigning shares and hope you can give some info here to help me determine if I should bring this up as a bug or feature request.

 

I've created a RAIDZ2 pool where I can make shares that get Exclusive access by default - Exclusive set in general preferences and the files are only accessed in the one pool, no secondary storage.

The limitation I'm seeing is when reassigning the location/pool for an existing share that is currently on the traditional array. Such a share has Exclusive = NO. That doesn't change to Yes when I reassign its location to the ZFS pool and set secondary storage to none. By definition it has the same settings as a new share (one pool only).

Can this limitation be remedied such that we gain exclusive access when making this change?

Does the directory still show on array drives?