Kernel Panic when disconnecting USB hardware

[hydrosIII]

Re: Kernel Panic when disconnecting keyboard

Just for curiosity, but in the meantime waiting the bug to be solved, why not just keep the keyboard connected or don't use keyboard at all (I always log in my servers with ssh)?

Is not the keyboard, maybe i will just modify the topic, is every usb. I mean i use ssh, the keyboard is just the way i figure out this bug, because i was working in the server with keyboard and monitor, because ssh was out. But i have ( had for now) a printer an a USB ups connected to the server. Everytime i turn the printer off, the server paniqued. Everytime the ups lose power the server paniqued. So that is why it is an issue. I could just switch the printer to another computer but the the unavoidable thing is the UPS. I need the ups plugged to the server.

 

[hydrosIII]

Re: Kernel Panic when disconnecting keyboard

Can a moderator change the topic of the thread to "Kernel Panic when disconnecting usb hardware" that would reflect better the problem. i Couldnt find the way to edit it myself.

Edit: Nervermind i just figured out.

 

[masterdam79]

I can confirm the same behaviour when switching my KVM between my two cluster nodes which effectively plugs and unplugs the USB and VGA ports

Verstuurd vanaf mijn Nexus 5 met Tapatalk

 

[hydrosIII]

I have oppened another bug report in the openvz bugzilla, since this seems to be a bug in the upstream kernel. Bug report is here: https://bugzilla.openvz.org/show_bug.cgi?id=2956

 

[hydrosIII]

I have oppened another bug report in the openvz bugzilla, since this seems to be a bug in the upstream kernel. Bug report is here: https://bugzilla.openvz.org/show_bug.cgi?id=2956

The openvz team just sent a reply on this bug. I have to provide a kernel crash dump of this issue. Any help in how to get that?? Can anyone provide a kernel crash dump to submit??.However their opinion is that there is a bug on the original Red Hat kernel. I will try to test this issue further however the issue is getting complicated. If anyone is interested in help me test. The help would be appreciated. I will try to install centos kernel in debian and test. Any helps of the proxmox devs?

 

[hydrosIII]

I have tried to install Centos kernel in debian and failed. I was just converting rpm packages with alien, i couldnt make any progress. I will try again later. Looking at my image of the kernel crash it seems to be provoqued by the colord-sane process. Maybe this is related to the panic. Has anyone experiencing these issues are running colord_sane?? Can anyone provide kernel crash dumps, so i can submit them to the bug report?

Also i have created another openvz node with another old machine to test if my hardware was bad, migrated all my vm to it, and the kernel panics are repeating so it is not hardware related.

 

[snowman66]

syslog message:

Jul 13 08:31:50 pve1 kernel: hub 1-1.3.4:1.0: port 1 disabled by hub (EMI?), re-enabling...

http://www.linuxquestions.org/quest...rt-nn-disabled-by-hub-emi-re-enabling-841593/

and this triggers kernel panic (at least in my case).

 

[jacques]

I am also experiencing this on a couple of hosts with kernel 2.6.32-31-pve running NUT (usbhid-ups) and it is very similar to Debian bug #634023.