2 hours ago, limetech said:

 

Any custom changes to your XML file?

No custom changes at all. bog standard xml created from the unraid VM manager page.

Has anyone had any luck with an OSX VM on RC15? I try rebooting the machine from the clover bootloader and smashing the DEL key to get into the UEFI/BIOS settings to change the resolution but it never enters the settings screen. It just goes right back to the bootloader.

 

Edit: Found somewhere else that F2 is what you need to press. Sorry!

Edited December 12, 20178 yr by ChadwickTheCrab

i have no issues with my high sierra vm. what version of clover are you using?

1 minute ago, david279 said:

i have no issues with my high sierra vm. what version of clover are you using?

I just figured out that it's no longer the delete key to change the OVMF settings, but now it's F2. I'm going to continue and attempt to install.

For you guys with VM issues, especially if you see this message:

 

[Firmware Bug]: TSC_DEADLINE disabled due to Errata; please update microcode to version: 0x3a (or later)

 

Please check if you are running the latest motherboard bios.

40 minutes ago, limetech said:

For you guys with VM issues, especially if you see this message:

 

[Firmware Bug]: TSC_DEADLINE disabled due to Errata; please update microcode to version: 0x3a (or later)

 

Please check if you are running the latest motherboard bios.

I have this message, but I have the latest motherboard BIOS.  Unfortunately, I have an older motherboard and I don't think ASROCK will be offering any more updates.

 

I get the same log entry in rc14:

[Firmware Bug]: TSC_DEADLINE disabled due to Errata; please update microcode to version: 0x22 (or later)

 

Edited December 12, 20178 yr by dlandon

24 minutes ago, limetech said:

For you guys with VM issues, especially if you see this message:

 

[Firmware Bug]: TSC_DEADLINE disabled due to Errata; please update microcode to version: 0x3a (or later)

 

Please check if you are running the latest motherboard bios.

Does it make any difference if we boot UEFI?

This is the only firmware bug that shows on my SuperMicro X11DPi-NT running latest bios. Not sure if this is relevant to the VM issue.

 

Dec 12 12:21:40 ServerNAS kernel: ACPI: [Firmware Bug]: BIOS _OSI(Linux) query ignored

I have this message too (and call traces), but all my VMs appear to work just fine.

[Firmware Bug]: TSC_DEADLINE disabled due to Errata; please update microcode to version: 0xb000020 (or later)

I do have the latest BIOS version installed (september 2017).

My Windows 7 Pro vm just hung on a restart. Force stop returns with Execution error. Failed to terminate process 32574 with SIGKILL: Device or resource busy.

 

Previously the same had happened to my Windows 10 vm as well but I put that down to just being Windows. 

Shutting down via the GUI that time resulted in unRaid seeing an unclean shutdown with the expected parity check on reboot.

 

No manual edits to either xml file.

 

Latest bios is installed.

 

ps ax give this for the process.

32574 ?        Zl    13:11 [qemu-system-x86] <defunct>

32581 ?        S      0:00 [vhost-32574]

32586 ?        S      0:00 [kvm-pit/32574]

 

Attached diagnostics are before I try to reboot again.

 

Kevin

tower-diagnostics-20171213-1111.zip

Windows 7 xml.txt

Windows 10 xml.txt

 

[Edit]

Rolled back to rc14. Shutdown via GUI. On poweron a parity check was started. Both vm's started, restarted and stopped with no errors.

I've added diagnostics taken after the poweron in the hope that they might show why an unclean shutdown happened.

 

 

 

tower-diagnostics-20171213-1250.zip

Edited December 13, 20178 yr by digitalfixer
Additional information

I"m having a strange issue with rc14 (not a typo). when I try to stop the array its almost as if it gets hung up? its having trouble stopping the array.

 

it hangs up on "Array Stopping•Retry unmounting user share(s)..."

cold-rewind-diagnostics-20171213-0809.zip

 

 

EDIT: Nevermind. I'm an idiot. I had left an SSH session open and was in one of those directories.. I just killed the session and the umount went fine. Excuse me while I hang my head in shame.

Edited December 13, 20178 yr by Darksurf

Most common cause of that is having a SSH session open, and the current directory is set to somewhere within /mnt/user

It looks like this is the error that Windows 10 says caused the VM to shutdown/hang:

Been seeing similar issues in the log since upgrading to 15e.  Typically happens when reopening a browser session upon reboot of a windows PC.

 

Dec 13 11:28:14 DaVault nginx: 2017/12/13 11:28:14 [error] 9634#9634: *1152195 limiting requests, excess: 5.646 by zone "authlimit", client: 192.168.1.168, server: , request: "POST /plugins/advanced.buttons/AdvancedButtons.php HTTP/1.1", host: "davault.local", referrer: "http://davault.local/Main"
Dec 13 11:28:14 DaVault nginx: 2017/12/13 11:28:14 [error] 9634#9634: *1152195 limiting requests, excess: 5.544 by zone "authlimit", client: 192.168.1.168, server: , request: "POST /plugins/advanced.buttons/AdvancedButtons.php HTTP/1.1", host: "davault.local", referrer: "http://davault.local/Main"
Dec 13 11:28:14 DaVault nginx: 2017/12/13 11:28:14 [error] 9634#9634: *1152195 limiting requests, excess: 5.494 by zone "authlimit", client: 192.168.1.168, server: , request: "GET /Main HTTP/1.1", host: "davault.local", referrer: "http://davault.local/Main"
Dec 13 11:28:14 DaVault nginx: 2017/12/13 11:28:14 [error] 9634#9634: *1152195 limiting requests, excess: 5.486 by zone "authlimit", client: 192.168.1.168, server: , request: "GET /favicon.ico HTTP/1.1", host: "davault.local", referrer: "http://davault.local/Main"

 

Also noticed blank pop ups during docker or plug-in upgrades.  Until the docker or plug-in page is manually refreshed, the blank message doesn't go away nor does it update when the upgrade process is finished.

 

6 minutes ago, luisv said:

Been seeing similar issues in the log since upgrading to 15e.  Typically happens when reopening a browser session upon reboot of a windows PC.

 

Dec 13 11:28:14 DaVault nginx: 2017/12/13 11:28:14 [error] 9634#9634: *1152195 limiting requests, excess: 5.646 by zone "authlimit", client: 192.168.1.168, server: , request: "POST /plugins/advanced.buttons/AdvancedButtons.php HTTP/1.1", host: "davault.local", referrer: "http://davault.local/Main"
Dec 13 11:28:14 DaVault nginx: 2017/12/13 11:28:14 [error] 9634#9634: *1152195 limiting requests, excess: 5.544 by zone "authlimit", client: 192.168.1.168, server: , request: "POST /plugins/advanced.buttons/AdvancedButtons.php HTTP/1.1", host: "davault.local", referrer: "http://davault.local/Main"
Dec 13 11:28:14 DaVault nginx: 2017/12/13 11:28:14 [error] 9634#9634: *1152195 limiting requests, excess: 5.494 by zone "authlimit", client: 192.168.1.168, server: , request: "GET /Main HTTP/1.1", host: "davault.local", referrer: "http://davault.local/Main"
Dec 13 11:28:14 DaVault nginx: 2017/12/13 11:28:14 [error] 9634#9634: *1152195 limiting requests, excess: 5.486 by zone "authlimit", client: 192.168.1.168, server: , request: "GET /favicon.ico HTTP/1.1", host: "davault.local", referrer: "http://davault.local/Main"

 

Also noticed blank pop ups during docker or plug-in upgrades.  Until the docker or plug-in page is manually refreshed, the blank message doesn't go away nor does it update when the upgrade process is finished.

 

 

 

That is happening because we enabled nginx rate limit for login.  That is, if you have set a root password then first access to webGui will present login dialog.  After you enter valid credentials the webGui continues loading.  This is to prevent brute-force attack on your password.  Probably we need to tune this a little differently.

15 minutes ago, luisv said:

Also noticed blank pop ups during docker or plug-in upgrades.  Until the docker or plug-in page is manually refreshed, the blank message doesn't go away nor does it update when the upgrade process is finished.

This is the "Advanced Buttons" plugin you've got installed doing this.  Nothing to do with 6.4 itself.  Alternatives is to setup CA AutoUpdate, or uninstall Advanced Buttons and do each update manually.

Edited December 13, 20178 yr by Squid

On 12/12/2017 at 4:17 PM, johnnie.black said:

Not all windows VMs, I have no problem with mine, and I run 3 (Win8.1, Win10 and Win Server 2012R2)

Like I mentioned I have no issues shutting down my VMs on rc15, but they have been crashing semi-regularly, mostly the Win8 and Win10 VMs, though this started before rc15, maybe rc12/13 if I had to guess, they never crashed before without an apparent reason.

 

I don't see nothing on the logs but diagnostics attached anyway, latest crash was the Win8 VM today at around 18:00H

tower1-diagnostics-20171213-1823.zip

1 hour ago, limetech said:

 

 

That is happening because we enabled nginx rate limit for login.  That is, if you have set a root password then first access to webGui will present login dialog.  After you enter valid credentials the webGui continues loading.  This is to prevent brute-force attack on your password.  Probably we need to tune this a little differently.

 

Thanks for explaining, and yes, password is set and exactly what happens upon browser refresh... it provides a logon prompt.  

 

1 hour ago, Squid said:

This is the "Advanced Buttons" plugin you've got installed doing this.  Nothing to do with 6.4 itself.  Alternatives is to setup CA AutoUpdate, or uninstall Advanced Buttons and do each update manually.

 

CA AutoUpdate is enabled, so I uninstalled Advanced Buttons.   

It appears the firmware bug issue is related to cpu microcode not being updated.  The command:

dmesg | grep -i microcode

Produces the output on my server:

[Firmware Bug]: TSC_DEADLINE disabled due to Errata; please update microcode to version: 0x22 (or later)

If this firmware bug is the cause of VM hang ups/not starting/failing then it may be that the Slackware microcode update package needs to be added.

Just to provide an update to everyone with VM issues, we did track down a bug report that is affecting the 4.14 Linux kernel related to stability issues with VMs (particular when attempting to shut down or reboot the VMs).  We will have more information on this soon.  Please stay tuned.

DL120 G6 now showing this:

 

Dec 14 11:02:40 Tower kernel: TCP: br0: Driver has suspect GRO implementation, TCP performance may be compromised.

Did not appear in 6.2.4

1 hour ago, 1812 said:

DL120 G6 now showing this:

 

Dec 14 11:02:40 Tower kernel: TCP: br0: Driver has suspect GRO implementation, TCP performance may be compromised.

Did not appear in 6.2.4

 

If you google that it appears this is probably a warning message introduced by some programmer who thinks this feature of the driver should be implemented.  I've noticed more of that happening in the kernel (nag messages telling developers to update their code).  Probably this is harmless.

Hi, im one of the lucky ones who doesnt have any VM issues actually with rc15, but what i see is, my disks are way more up then before ...

 

where could i track down what is causing the disk accesses ?

54 minutes ago, alturismo said:

where could i track down what is causing the disk accesses ?

Dynamix Active Streams plugin

dlandon's Open Files plugin

17 minutes ago, Squid said:

Dynamix Active Streams plugin

dlandon's Open Files plugin

 

thanks alot