so,
cause eBPF pulls all from the kernel,
yes you can.
but you need to dive into the stuff..
Knowing basics about performance and kernel can help.
but it's possible.

slainte!


26.12.2022 19:19:47 Gk Gk <ygk.kmr@gmail.com>:

So, can I derive cpu, memory, disk and network stats of the VMs ?

On Monday, December 26, 2022, Marko Horn <weber@zbfmail.de> wrote:
> Hello,
>
> yes I can :-)
>
> https://www.brendangregg.com/
>
> on the page you will find a link to his 'Linux Performance' book.
> On the page are also further Infos about the topic 'performance'.
> you will have to dive into the stuff.
> (e)BPF is a "MIGHTY" area.
> But you can pull out nearly 'every' meta data from the kernel.
>
> Also for your interest:
>
> https://www.brendangregg.com/Slides/SREcon2022_ComputingPerformance/
>
> Computing Performance 2022 and what is on the horizon!
>
> have fun!
>
> Slainte
>
> ________________________________
>
> 26.12.2022 18:09:39 Gk Gk <ygk.kmr@gmail.com>:
>
> Thanks Marco for the response. Is it possible to collect these vm stats from the host KVM hypervisor ? Also can u give the link for this book ?
>
> Thanks
> Happy holidays
>
> On Monday, December 26, 2022, Marko Horn <weber@zbfmail.de> wrote:
>> Hi,
>>
>> I think best way is to use BPF to collect stats. I did this in past in combination with 'perf'.
>> Theres a nice book of Brandon Greg from Netflix Networks where he describe how to do this and how to write own scripts.
>>
>> merry Xmas
>>
>> marko
>>
>> ________________________________
>>
>> 26.12.2022 11:34:20 Gk Gk <ygk.kmr@gmail.com>:
>>
>> Hi All,
>> I am trying to collect memory, disk and network stats for a VM on kvm host. It seems that the statistics are not matching what the OS inside the VM is reporting. Why is this discrepancy ?
>> Is this a known bug of libvirt ? Also I heard that libvirt shows cumulative figures for these measures ever since the VM was created.  Also I tested by creating a new vm and comparing the stats without a reboot . Even in this case, the stats dont agree.  Can someone help me here please ?
>>
>> Thanks
>> Kumar
>>