Eventually it "just" executes commands in both cases. To do so (other than my shell script test) it does that through virCommandRun which always runs async by virCommandRunAsync and then waits with virCommandWait - their implementation did not change either.
But I was setting up an strace to check the exact timings of those calls and that not only took longer (overhead - ok), but the zesty version took much longer. I don't see yet why, need to read all the logs gathered (hundreds of megabytes). While doing so I found that libvirt seems really 100% busy on one core. That might be an artifact of the stracing, but if not this could be analyzed with a profiler like perf. But that would need VMs to run that in, while currently I have containers to use the same kernel in both cases (to eliminate this as the source). -- You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. https://bugs.launchpad.net/bugs/1727366 Title: virsh start/destroy is too slow after adding firewall rule To manage notifications about this bug go to: https://bugs.launchpad.net/ubuntu/+source/libvirt/+bug/1727366/+subscriptions -- ubuntu-bugs mailing list [email protected] https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
