| 1 |
Am 2018-05-30 um 10:46 schrieb Bill Kenworthy: |
| 2 |
> On 30/05/18 15:52, Stefan G. Weichinger wrote: |
| 3 |
>> maybe someone has hit that as well: |
| 4 |
>> |
| 5 |
>> https://bugs.gentoo.org/656886 |
| 6 |
>> |
| 7 |
>> bug report at redhat is from yesterday (!) |
| 8 |
>> |
| 9 |
>> - |
| 10 |
>> |
| 11 |
>> That windows guest is shutting down and/or crashing now and then. |
| 12 |
>> Same QEMU/libvirt combo at another site runs >20 VMs without problems. |
| 13 |
>> |
| 14 |
>> I'd be happy to find some workaround or fix as it is a bit problematic |
| 15 |
>> to have the customer email every few days that his server VM is down. |
| 16 |
>> Gives bad reputation ... |
| 17 |
>> |
| 18 |
>> If helpful I will post VM.xml etc |
| 19 |
>> |
| 20 |
>> Stefan |
| 21 |
>> |
| 22 |
> error messages? - I get this happening sometimes from the oom killer |
| 23 |
> when the server runs out of memory - semi-random (usually but not |
| 24 |
> always the same vm) |
| 25 |
|
| 26 |
I mentioned the libvirt error message in the bug report. |
| 27 |
What else is interesting? |
| 28 |
|
| 29 |
The gentoo host has 16 GB RAM, and the single VM is configured with 8 |
| 30 |
gigs ... so that looks ok to me. |
| 31 |
|
| 32 |
host # free -m |
| 33 |
total used free shared buff/cache |
| 34 |
available |
| 35 |
Mem: 16015 8496 3138 5 4379 |
| 36 |
7261 |
| 37 |
Swap: 8191 10 8181 |
Archives