unRAID Server Release 6.0-rc3-x86_64 Available


447 posts in this topic Last Reply

Recommended Posts

Also something else,

 

in the VM manager if you click stop VM you have to refresh the page before it shows the VM as stopped, tried with the display updates set to regular and set to realtime, same with both settings.

 

clicking start on a vm and the icon refreshes, click force stop and it refreshes, however clicking stop it doesn't unless you physically refresh the page.

 

 

Try waiting longer. Eric converted the page to ajax and it will auto update the VM icon / status when the shutdown completes.

 

On a side note. I've asked Eric in a future build to adjust how this works so that when you tell the VM to shutdown, the icon changes to show a shutdown in progress.

 

start the clock, lol.

Link to post
  • Replies 446
  • Created
  • Last Reply

Top Posters In This Topic

went for a full 6 minutes, it doesn't show stopped until you refresh the page.

 

i have regular set as page refresh, also tried real time , makes no difference, the VM indicator shows green until the page is refreshed.

 

GngMAF0.png

 

 

you can see the counter thing in this image.

Link to post

Is there a reason i can not get VM to function,  was working in beta 15 but i didnt have a use for it then so just played with a windows VM then removed it.

 

now when i goto Settings -> VM Manager to re-enable it as i now have a need for one  i get this

 

Your hardware does not have Intel VT-x or AMD-V capability. This is required to create VMs in KVM. Click here to see the unRAID Wiki for more information

 

But i know it was working in beta 15 and is enabled in dmesg

 

root@Tower:~# dmesg | grep AMD-Vi

AMD-Vi: Found IOMMU at 0000:00:00.2 cap 0x40

AMD-Vi: Interrupt remapping enabled

AMD-Vi: Lazy IO/TLB flushing enabled

 

 

am i missing something or is this what  eschultz was referring to

 

Regex parsing failure in the code for mainly AMD devices, I found and fixed the bug.

 

Do you get any errors back if you try to execute this command from the console?:

 

modprobe kvm_amd

 

 

Could not figure it out so i rolled back to beta 15,  did the settings page stuff,  then updated to RC1 and have no problems managing VM's 

Link to post

went for a full 6 minutes, it doesn't show stopped until you refresh the page.

What browser / version?

 

firefox 37.0.2 on mac osx 10.10.3

The one browser I didn't test. Can you try safari or chrome just to confirm if those work for you?

Link to post

Updated both pro servers from previous beta. Was greeted with 5 drive slots and all drives empty on both of them. I freaked out but then noticed Slots was set to 5 even though I'm on Pro. I set that to 24, and it automatically re-assigned all the drives.

 

Just a heads up.

 

I experienced the same issue after downloading and updating.

I did the reboot and the system had no assigned drives, was set to 5 and was stopped.

I changed the drive count to 24 as I am also running Pro.

Resetting the drive count enabled the Start button and the system assigned the corrrect number of drives I have installed which is 11 and the system started.

A bit disconcerting and should be spelled out in the OP first page or corrected.

Didn't happen to me, but 5 is the correct number of data drives for me. I also have Pro and 2 cache slots.
Link to post

went for a full 6 minutes, it doesn't show stopped until you refresh the page.

What browser / version?

 

firefox 37.0.2 on mac osx 10.10.3

The one browser I didn't test. Can you try safari or chrome just to confirm if those work for you?

 

can test safari, don't have chrome.

Link to post

same with safari, doesn't show stopped until the page is refreshed.

Eric confirmed for me that the ajax events arent totally complete yet. My bad.

Link to post

Updated both pro servers from previous beta. Was greeted with 5 drive slots and all drives empty on both of them. I freaked out but then noticed Slots was set to 5 even though I'm on Pro. I set that to 24, and it automatically re-assigned all the drives.

 

Talk about a heart-stop!  Thanks for/to your email; I'm back up and running.  Thanks again for posting!

Link to post

OK I thought I was having a browser cache issue as well. Why? Because I just went back into VM Manger and it was working despite making no changes or reboots for 2 hours.  But then I remembered (because the PuTTY window was still open) that Eric had me issue a modeprobe kvm-amd command!

 

So I rebooted, and I once again show,

Your hardware does not have Intel VT-x or AMD-V capability. This is required to create VMs in KVM.

Then I issued modprobe kvm_amd at a command line and BAM, VM Manager is functional!!!!!!!!!!!!!!!!!

 

So, the module is NOT being loaded on its own.

 

But I also still have to report: [glow=red,2,300]system will NOT boot at all without iommu=pt appended to syslinux[/glow]; and iommu=1 might be needed for full VM functionality but I have not dug into that yet. I'm not sure what is missing from the kernel by default, but it should probably be added to the install notes until fixed.

 

A note about the modprobe kvm_amd issue might be advisable. This one is very repeatable. I just did it again for fun. Reboot, VM Manager fail, issue modprobe kvm_amd, refresh VM Manager pade, SUCCESS, like clockwork.

 

Syslog is attached, and interestingly I do see failures of modeprob in the log with time stamps before I manually issue it. You can see when I logged in and then the messages change from no hardware support to already loaded:

May 16 14:28:29 Tower kernel: kvm_intel: unknown parameter 'kvm_amd' ignored
May 16 14:28:29 Tower kernel: kvm: no hardware support
May 16 14:28:30 Tower kernel: kvm_intel: unknown parameter 'kvm_amd' ignored
May 16 14:28:30 Tower kernel: kvm: no hardware support
May 16 14:28:31 Tower kernel: kvm_intel: unknown parameter 'kvm_amd' ignored
May 16 14:28:31 Tower kernel: kvm: no hardware support
May 16 14:28:38 Tower sshd[3931]: Accepted password for root from 10.0.0.6 port 17100 ssh2
May 16 14:28:42 Tower kernel: kvm: Nested Virtualization enabled
May 16 14:28:42 Tower kernel: kvm: Nested Paging enabled
May 16 14:28:44 Tower kernel: kvm_intel: unknown parameter 'kvm_amd' ignored
May 16 14:28:44 Tower kernel: kvm: already loaded the other module
May 16 14:31:42 Tower kernel: kvm_intel: unknown parameter 'kvm_amd' ignored
May 16 14:31:42 Tower kernel: kvm: already loaded the other module

syslog.zip

Link to post

Updated both pro servers from previous beta. Was greeted with 5 drive slots and all drives empty on both of them. I freaked out but then noticed Slots was set to 5 even though I'm on Pro. I set that to 24, and it automatically re-assigned all the drives.

 

Talk about a heart-stop!  Thanks for/to your email; I'm back up and running.  Thanks again for posting!

 

Updated the announcement post with explanation and workaround for this issue.

Link to post

OK I thought I was having a browser cache issue as well. Why? Because I just went back into VM Manger and it was working despite making no changes or reboots for 2 hours.  But then I remembered (because the PuTTY window was still open) that Eric had me issue a modeprobe kvm-amd command!

 

So I rebooted, and I once again show,

Your hardware does not have Intel VT-x or AMD-V capability. This is required to create VMs in KVM.

Then I issued modprobe kvm_amd at a command line and BAM, VM Manager is functional!!!!!!!!!!!!!!!!!

 

So, the module is NOT being loaded on its own.

 

But I also still have to report: [glow=red,2,300]system will NOT boot at all without iommu=pt appended to syslinux[/glow]; and iommu=1 might be needed for full VM functionality but I have not dug into that yet. I'm not sure what is missing from the kernel by default, but it should probably be added to the install notes until fixed.

 

A note about the modprobe kvm_amd issue might be advisable. This one is very repeatable. I just did it again for fun. Reboot, VM Manager fail, issue modprobe kvm_amd, refresh VM Manager pade, SUCCESS, like clockwork.

 

Syslog is attached, and interestingly I do see failures of modeprob in the log with time stamps before I manually issue it. You can see when I logged in and then the messages change from no hardware support to already loaded:

May 16 14:28:29 Tower kernel: kvm_intel: unknown parameter 'kvm_amd' ignored
May 16 14:28:29 Tower kernel: kvm: no hardware support
May 16 14:28:30 Tower kernel: kvm_intel: unknown parameter 'kvm_amd' ignored
May 16 14:28:30 Tower kernel: kvm: no hardware support
May 16 14:28:31 Tower kernel: kvm_intel: unknown parameter 'kvm_amd' ignored
May 16 14:28:31 Tower kernel: kvm: no hardware support
May 16 14:28:38 Tower sshd[3931]: Accepted password for root from 10.0.0.6 port 17100 ssh2
May 16 14:28:42 Tower kernel: kvm: Nested Virtualization enabled
May 16 14:28:42 Tower kernel: kvm: Nested Paging enabled
May 16 14:28:44 Tower kernel: kvm_intel: unknown parameter 'kvm_amd' ignored
May 16 14:28:44 Tower kernel: kvm: already loaded the other module
May 16 14:31:42 Tower kernel: kvm_intel: unknown parameter 'kvm_amd' ignored
May 16 14:31:42 Tower kernel: kvm: already loaded the other module

Thanks for the detailed info -- I think I have a good idea on how to fix this issue now.  I should probably invest in a new AMD cpu for testing... I got 99 computers but AMD ain't one

Link to post

[spit] hahaha well i'm glad i was actually able to help vice just complaining. Thanks for that :)

 

In the mean time I've added modprobe to my GO file, but haven't rebooted. I'm guessing it would work. I don't reboot much except for unraid updates so either way I think I'm good for now. Who knows, maybe I'll get around to rebuilding my Arch VM in KVM now.

 

Cheers

Link to post

Just updated to rc-1. After updating, my graphics card is no longer showing as an available device to assign as a graphics card or sound card when updating a VM or trying to create a new VM. It was passing through fine before the upgrade.

 

The only choice I have for Graphics Cards is VNC when editing or adding a VM.  I have a iGPU and a pass through AMD graphics card.  Neither one shows as a choice.

 

Regex parsing failure in the code for mainly AMD devices, I found and fixed the bug.

 

iGPUs are now hidden by design in rc1 -- technically these won't pass-through, probably ever.

 

Hi. Same issue for me passing through my AMD card (HD 6450).  Glad to hear the bug is now fixed but any idea when we might see that fix rolled out?  I might want to roll back to the beta if it's likely to be a while.  Thanks.

Link to post

I've lost my disk configuration after upgrading to the RC1 and the disk slots are all showing up as unassigned. 

 

Of course I didn't make a backup or screenshot my disk config before doing the upgrade.  Bad me.

 

I've checked the backup folders for beta15 and the config files aren't there.

 

How can I get my disk config back properly?

 

EDIT:  I changed the number of disk slots to 7 and the disks are now showing up.  Configuration is showing as valid now as well.  Going to start the array and see how it goes.

 

 

 

 

Link to post

I've lost my disk configuration after upgrading to the RC1 and the disk slots are all showing up as unassigned. 

 

Of course I didn't make a backup or screenshot my disk config before doing the upgrade.  Bad me.

 

I've checked the backup folders for beta15 and the config files aren't there.

 

How can I get my disk config back properly?

 

EDIT:  I changed the number of disk slots to 7 and the disks are now showing up.  Configuration is showing as valid now as well.  Going to start the array and see how it goes.

Already reported. See addendum to first post.
Link to post

Has anyone else had these messages when mover is run?

 

Cannot stat file /proc/12396/fd/26: No such file or directory

 

 

May 17 08:50:01 server-ramford kernel: mdcmd (60): spindown 5
May 17 08:50:08 server-ramford kernel: mdcmd (61): spindown 4
May 17 08:52:13 server-ramford kernel: mdcmd (62): spindown 1
May 17 08:57:45 server-ramford kernel: mdcmd (63): spindown 0
May 17 09:01:14 server-ramford emhttp: read_line: client closed the connection
May 17 09:01:44 server-ramford emhttp: shcmd (43747): /usr/local/sbin/mover |& logger &
May 17 09:01:44 server-ramford logger: mover started
May 17 09:01:44 server-ramford logger: moving "TV"
May 17 09:01:44 server-ramford logger: Cannot stat file /proc/12396/fd/26: No such file or directory
May 17 09:01:44 server-ramford logger: Cannot stat file /proc/12396/fd/36: No such file or directory
May 17 09:01:44 server-ramford logger: Cannot stat file /proc/12399/fd/11: No such file or directory
May 17 09:01:44 server-ramford logger: Cannot stat file /proc/12399/fd/31: No such file or directory
May 17 09:01:44 server-ramford logger: ./TV/ZZZ - S02E01 - SDTV.mp4
May 17 09:01:44 server-ramford logger: .d..t...... ./
May 17 09:01:44 server-ramford logger: .d..t...... TV/
May 17 09:01:44 server-ramford logger: .d..t...... TV/ZZZ/
May 17 09:01:44 server-ramford logger: .d..t...... TV/ZZZ/Season 2/
May 17 09:01:44 server-ramford logger: >f+++++++++ TV/ZZZ/Season 2/ZZZ - S02E01 - SDTV.mp4
May 17 09:01:48 server-ramford emhttp: /usr/bin/tail -n 42 -f /var/log/syslog 2>&1
May 17 09:01:50 server-ramford logger: ./TV/ZZZ/Season 2
May 17 09:01:50 server-ramford logger: .d..t...... TV/ZZZ/Season 2/
May 17 09:01:50 server-ramford logger: ./TV/ZZZ
May 17 09:01:50 server-ramford logger: .d..t...... TV/ZZZ/
May 17 09:01:50 server-ramford logger: ./TV/
May 17 09:01:50 server-ramford logger: .d..t...... TV/
May 17 09:01:50 server-ramford logger: skipping "appdata"
May 17 09:01:50 server-ramford logger: skipping "apps"
May 17 09:01:50 server-ramford logger: mover finished

Link to post
Guest
This topic is now closed to further replies.