other: 0.978
debug: 0.971
PID: 0.966
permissions: 0.960
device: 0.957
semantic: 0.957
graphic: 0.956
KVM: 0.936
performance: 0.933
socket: 0.925
files: 0.918
boot: 0.906
network: 0.884
vnc: 0.864

migrate:  add tls option in virsh, migrate failed

version:
    libvirt-3.4.0 + qemu-2.9.90(latest)

domain:
    any

step:
    1. generate tls certificate in /etc/pki/libvirt-migrate
    2. start vm
    3. migrate vm, cmdline:
       virsh migrate rh7.1-3 --live --undefinesource --persistent --verbose --tls qemu+ssh://IP/system
    4. then migrate failed and reported:
 Migration: [ 64 %]error: internal error: qemu unexpectedly closed the monitor: Domain pid=5288, libvirtd pid=49634
 kvm: warning: CPU(s) not present in any NUMA nodes: 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
 kvm: warning: All CPU(s) up to maxcpus should be described in NUMA config
 kvm: error while loading state section id 2(ram)
 kvm: load of migration failed: Input/output error

other:
    Analysis qemu code and debug:
#0  ram_save_page (f=0x55ca8be370e0, pss=0x7fefdfc7b9a0, last_stage=false, bytes_transferred=0x55ca885ec1d8) 
#1  0x000055ca87b00b21 in ram_save_target_page (ms=0x55ca885b8d80, f=0x55ca8be370e0, pss=0x7fefdfc7b9a0, last_stage=false, bytes_transferred=0x55ca885ec1d8, dirty_ram_abs=0)
#2  0x000055ca87b00bda in ram_save_host_page (ms=0x55ca885b8d80, f=0x55ca8be370e0, pss=0x7fefdfc7b9a0, last_stage=false, bytes_transferred=0x55ca885ec1d8, dirty_ram_abs=0)
#3  0x000055ca87b00d39 in ram_find_and_save_block (f=0x55ca8be370e0, last_stage=false, bytes_transferred=0x55ca885ec1d8)
#4  0x000055ca87b020b8 in ram_save_iterate (f=0x55ca8be370e0, opaque=0x0)
#5  0x000055ca87b07a9a in qemu_savevm_state_iterate (f=0x55ca8be370e0, postcopy=false) 
#6  0x000055ca87e404e5 in migration_thread (opaque=0x55ca885b8d80)

This is the qemu bug tracker here. Please report libvirt bugs to the libvirt project instead (see http://libvirt.org/bugs.html). Thanks!

with libvirt test, qemu report error, this is problem for qemu，so report here!

But then please follow this recommendation from http://www.qemu.org/contribute/report-a-bug/ :

"Reproduce the problem directly with a QEMU command-line. Avoid frontends and management stacks, to ensure that the bug is in QEMU itself and not in a frontend."

So can you reproduce the problem by starting QEMU directly from the command line?

ok, thank you!

chan: The backtrace you show there - how did you capture that? Was that from a core dump  ?  If so then please do a   bt full  to get more detail.

Even if you can't reproduce it with qemu without libvirt I'd be interested to debug it - but the libvirt logs from /var/log/libvirt/qemu/.... on both the source and the destination would help.


[Expired for QEMU because there has been no activity for 60 days.]