Determining when a guest booted / how long it's been running

[Digimer]

On 2020-10-01 1:21 a.m., Michal Prívozník wrote:
> On 10/1/20 3:42 AM, Digimer wrote:
>> Hi all,
>>
>>    Is there a way to tell when a tool made a change to guest (ie: used
>> virt-manager to make a change)? Following, is there a way to check to
>> see if there are changes queued to take effect when the guest next
>> reboots?
>>
> 
> You can listen for events. For changes to inactive XML you will get
> VIR_DOMAIN_EVENT_DEFINED+VIR_DOMAIN_EVENT_DEFINED_UPDATED lifecycle
> event (+reason) and for changes to live XML (like hotplug and hotunplug)
> you'll get DEVICE_ADDED or DEVICE_REMOVED.
> 
> But there is no API/virsh command that you could call to get the
> timestamp of last modification. You have to listen for events.
> 
> And the second question - you can dump live and inactive XML and see if
> there is any difference (modulo runtime configuration from live XML).
> The live XML should be strictly bigger than inactive XML, meaning live
> XML should be inactive XML + runtime info. But again, no API there,
> because changes to inactive XML can be done any time - libvirt keeps
> live and inactive XMLs separately.
> 
>>    If either of the above are not possible, is there a way to see when
>> a guest last booted or how long a guest has been running?
> 
> Again, if listening to evens - libvirt emits one when a guest is
> started. But I don't think we store a timestamp of start anywhere nor
> expose it through an API. Maybe there is some indirect way?
> 
> Michal

Thanks, Michal.

Now I know not to keep looking. I do infer the boot time (explained in
my reply to daggs and feri. In brief, when we notice a server is
running, that had last been seen off, we use the pid to get the runtime
in seconds and subtract that from the current time to get a boot time.

-- 
Digimer
Papers and Projects: https://alteeve.com/w/
"I am, somehow, less interested in the weight and convolutions of
Einstein’s brain than in the near certainty that people of equal talent
have lived and died in cotton fields and sweatshops." - Stephen Jay Gould