[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xen-users] domU kernel crash after live migration on Debian 7.5

On Tue, Jan 27, 2015 at 4:33 PM, Ian Campbell <Ian.Campbell@xxxxxxxxxx> wrote:
Are these HVM or PV guests?

Are your physical machines all the same (e.g. processor)?Â

Actually I have a mix of the following two CPUs:

- Intel(R) Xeon(R) CPU E5-2650 v2 @ 2.60GHz
- Intel(R) Xeon(R) CPU E5645 @ 2.40GHz
Are you doing any cpuid levelling at all?Â

I don't know what cpuid levelling is, so I would just answer here: whatever is default in Xen.
> [Â 223.557790] invalid opcode: 0000 [#1] SMP
> [Â 223.557825] RIP: e030:[<ffffffff81243dbd>]Â [<ffffffff81243dbd>] arch_get_random_long+0x5/0x15

An interesting place to crash, not something I've ever seen before.

> [Â 223.558066] Code: 43 81 48 89 e9 48 89 df e8 13 e9 e8 ff 83 f8 01
>Â 19 d2 f7 d2 83 e2 f5 5b 5d 89 d0 41 5c c3 e8 d2 02 dd ff 66 90 c3 ba
> 0a 00 00 00 <48> 0f c7 f0 72 04 ff ca 75 f6 48 89 07 89 d0 c3 41 57

Interestingly "0f c7" (seen after the RIP at <48>) is the "rdrand"
instruction which is a newish instruction (sandybridge onwards). I
wonder if perhaps you've migrated to a machine which doesn't support it
for some reason (older processor, disabled in the bios, something else)?
That would certainly fit the symptoms.

My guess here would be the older processor which actually never popped up into my mind. I thought it would be possible to live migrate live between any type of processors as long as the architecture is the same. Do you have any pointers or documentation about how "much" processors are allowed to differ for live migration still being able to do its job correctly? For reference here is Intel's comparison of my two CPUs: http://ark.intel.com/compare/75269,48768


Xen-users mailing list



Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.