I am pretty new to XenServer so it may just be something silly causing this. Some background first. I have 2 instances of XenServer running on my network. I have an older horsepower server (quadsocket amd opterons 8224s, 64gb of ram) that doesnt support HVM, and a newer consumer grade system on a Gigabyte motherboard, with an AMD A6-6400k and 4gb of ram, that does support virtualization. The second machine was set up to do vmware to xen conversions, since my ESXi license does not allow XCM to export VMs. I am currently migrating from VMware ESXi to XenServer.

Now the horsepower server has been idle for a couple weeks with absolutely no issues. I just got my conversion server running a couple days ago. I noticed my conversion XenServer has been rebooting itself occasionally, with no warning. This morning, before I left for work, i noticed it was completely hung. No response from the local console either, just a blank screen on my monitor, and I had to restart the system to get XenServer to respond again.

I have looked around the net for a similar issue, and all I can find is a similar occurence on Nephalem chipset based Xeon processors. I neither use Xeons nor have any C-state options, that I am aware of, in my bios. Could it be UEFI or something causing my hangs?

I am not sure which logs would be useful to post here, and I dont see anything related to the hang up in my logs, only missing timestamps.

Any and all suggestions are greatly appreciated.

  • 1
    This appears to be a hardware problem and not a XenServer problem. And XS does not support UEFI. Use legacy mode to boot it. – Vinícius Ferrão Apr 22 '15 at 23:53
  • Run a memory tester, replace the bad DIMM when it tells you which one is junk. – yoonix Apr 23 '15 at 0:01
  • I appreciate the suggestions. I will try again in legacy when I get off work in a couple of hours. I ran a full hardware check before installing XenServer on the conversion server, mainly just to make sure the new 3x 2tb drives were good to go (this machine will also double as a fileserver once my conversions are done), so I doubt it is a bad DIMM, but I will double-check as well when I get home from work. – Vl4dim1r Apr 23 '15 at 2:20
up vote 0 down vote accepted

I just got home from work and checked out the Bios on my conversion server. Turns out this Gigabyte motherboard has C-states, only named "C6" instead, so I disabled both C-states and Turbo. So far so good. If anyone else has this issue, look for the "C6" and "CPB" options in "M.I.T. Information -> Advanced Frequency Settings" in your BIOS and disable those.

I ended up messing with some other options as well. Disabling UEFI somehow made my motherboard unable to boot. Luckily this motherboard has a dual BIOS option, so I just had to switch to my 2nd stock BIOS and set those C-state/Turbo options.

Reference: http://support.citrix.com/article/CTX127395 - This is the guide for disabling C-states. This was specifically written for Nephalem chipset based Xeon processors, but my AMD A6 server seems to be operating as intended after disabling C-states. All the C-states do is enable the BIOS to make on-the-fly adjustments to the processor's speed/performance. This would be handy to put out extra mhz when more work is requested from clients, but if it means my host will hang randomly I am okay without it.

Try installing XenServer Service Pack 1 (XS65SP1) and see if that fixes it.

Here are similar issues:

http://discussions.citrix.com/topic/364830-crashes-after-update-to-65/ http://discussions.citrix.com/topic/359694-random-reboots-after-upgrade-to-65/

Your Answer

 

By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.

Not the answer you're looking for? Browse other questions tagged or ask your own question.