> Is it really that big if you trace an otherwise idle VM?
That was probably from taking the last sample during the boot process. I
waited for the boot process to complete this time before tracing, which
yielded a much smaller file (9.8M).
Due to the delay in trace-cmd on this host, it's difficult for me to get
very short samples (~2sec). I hope that the additional sample data isn't
a problem.
Here's the post-boot/2sec/4sec experiment run again, with "trace-cmd
record -e kvm".
http://45.63.6.248/kvm3/trace-01-boot.tar.gz
http://45.63.6.248/kvm3/trace-02-2sec.tar.gz
http://45.63.6.248/kvm3/trace-03-4sec.tar.gz
http://45.63.6.248/kvm3/env.txt
Best,
Andrew
On 1/2/2018 12:39 PM, Paolo Bonzini wrote:
On 02/01/2018 18:34, Andrew Davis wrote:
trace-01-boot.tar.gz (16G -> 1.7G compressed), trace-cmd record -e kvm
I can tell it's this one from the size. :) Note that I only really need
a few seconds of trace (even 1-2 seconds until the ticks start slowing
down, but I have no idea what's happening after that point which is why
I said 10 originally).
Is it really that big if you trace an otherwise idle VM?
Paolo
trace-02-date-loop-started.tar.gz (4.9M -> 1M compressed), trace-cmd
record -e kvm -c -F -P 19582
trace-03-date-loop-running.tar.gz (5.3M -> 1.1M compressed) trace-cmd
record -e kvm -F -P 19582
http://45.63.6.248/kvm2/trace-01-boot.tar.gz
http://45.63.6.248/kvm2/trace-02-date-loop-started.tar.gz
http://45.63.6.248/kvm2/trace-03-date-loop-running.tar.gz
