VM's kernel gets panicked after a successful block live migration

asked 2016-10-10 00:05:32 -0600

fifi gravatar image

updated 2016-10-20 20:47:21 -0600

I have configured block live migration on Juno. Everything works fine. Sometimes, especially when the vm is loaded and its flavor is medium or larger, the block live migration finishes successfully but migrated vm is not responsive anymore (no ping, no ssh, no services). When I check the vm's console on horizon, I see this:

   [3958.698908] general protection fault: fff2 [#1] SMP 
[ 3958.698908] Modules linked in: ib_iser rdma_cm iw_cm ib_cm ib_sa ib_mad ib_core ib_addr iscsi_tcp libiscsi_tcp libiscsi scsi_transport_iscsi dm_crypt joydev serio_raw i2c_piix4 mac_hid hid_generic crct10dif_pclmul crc32_pclmul aesni_intel aes_x86_64 lrw gf128mul glue_helper ablk_helper cryptd syscopyarea sysfillrect sysimgblt usbhid ttm drm_kms_helper psmouse hid drm floppy pata_acpi
[ 3958.698908] CPU: 0 PID: 0 Comm: swapper/0 Not tainted 3.13.0-95-generic #142-Ubuntu
[ 3958.698908] Hardware name: OpenStack Foundation OpenStack Nova, BIOS Bochs 01/01/2011
[ 3958.698908] task: ffffffff81c15480 ti: ffffffff81c00000 task.ti: ffffffff81c00000
[ 3958.698908] RIP: 0010:[<ffffffff81051586>]  [<ffffffff81051586>] native_safe_halt+0x6/0x10
[ 3958.698908] RSP: 0018:ffffffff81c01eb8  EFLAGS: 00000286
[ 3958.698908] RAX: 00000000ffffffed RBX: ffffffffffffffff RCX: 0100000000000000
[ 3958.698908] RDX: 0000000000000000 RSI: 0000000000000000 RDI: 0000000000000046
[ 3958.698908] RBP: ffffffff81c01eb8 R08: 0000000000000000 R09: 0000000000000000
[ 3958.698908] R10: 0000000000000000 R11: 0000000000000000 R12: 0000000000000000
[ 3958.698908] R13: ffffffff81c01fd8 R14: ffffffff81de02c0 R15: ffffffff81c01fd8
[ 3958.698908] FS:  0000000000000000(0000) GS:ffff88023fc00000(0000) knlGS:0000000000000000
[ 3958.698908] CS:  0010 DS: 0000 ES: 0000 CR0: 000000008005003b
[ 3958.698908] CR2: 00007fc0c9909000 CR3: 0000000036200000 CR4: 00000000000006f0
[ 3958.698908] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[ 3958.698908] DR3: 0000000000000000 DR6: 0000000000000000 DR7: 0000000000000000
[ 3958.698908] Stack:
[ 3958.698908]  ffffffff81c01ed8 ffffffff8101db0f ffffffffffffffff ffffffff81d147e0
[ 3958.698908]  ffffffff81c01ee8 ffffffff8101e426 ffffffff81c01f38 ffffffff810c1cc1
[ 3958.698908]  ffffffff81c01fd8 ffffffff81c01fd8 6ea03e49a33e14fd ffffffffffffffff
[ 3958.698908] Call Trace:
[ 3958.698908]  [<ffffffff8101db0f>] default_idle+0x1f/0x100
[ 3958.698908]  [<ffffffff8101e426>] arch_cpu_idle+0x26/0x30
[ 3958.698908]  [<ffffffff810c1cc1>] cpu_startup_entry+0xc1/0x2b0
[ 3958.698908]  [<ffffffff81718047>] rest_init+0x77/0x80
[ 3958.698908]  [<ffffffff81d34f70>] start_kernel+0x438/0x443
[ 3958.698908]  [<ffffffff81d34941>] ? repair_env_string+0x5c/0x5c
[ 3958.698908]  [<ffffffff81d34120>] ? early_idt_handler_array+0x120/0x120
[ 3958.698908]  [<ffffffff81d345ee>] x86_64_start_reservations+0x2a/0x2c
[ 3958.698908]  [<ffffffff81d34733>] x86_64_start_kernel+0x143/0x152
[ 3958.698908] Code: 00 00 00 00 00 55 48 89 e5 fa 5d c3 66 0f 1f 84 00 00 00 00 00 55 48 89 e5 fb 5d c3 66 0f 1f 84 00 00 00 00 00 55 48 89 e5 fb f4 <5d> c3 0f 1f 84 00 00 00 00 00 55 48 89 e5 f4 5d c3 66 0f 1f 84 
[ 3958.698908] RIP  [<ffffffff81051586>] native_safe_halt+0x6/0x10
[ 3958.698908]  RSP <ffffffff81c01eb8>
[ 3958.843203] ---[ end trace 700f42850fcd2598 ]---
[ 3958.843603] Kernel panic - not syncing: Attempted to kill the idle task!

To make the migrated vm work on the new host, I need to soft reboot it. Does anyone knows what causes this?

By the Way, I use QEMU and Libvirt version 1.2.2 and the image I use is standard ubuntu 14.04 cloud image 64bit downloaded through "ubuntu clod image" website. Both compute nodes have identical hardware and Virtualization has been also activated on them.

Thanks

edit retag flag offensive close merge delete