Centos-6 x64 guest regularly crashes

[Victor Abramoff]

Hi,

My Centos-6 x64 guest regularly crashes after several days of uptime.

VirtualBox: 5.0.6 r103037 (same problem with many previous releases).
Host: Windows Server 2008 x64, with latest updates, 24 GB RAM.
Guest: Centos-6 x64, with latest updates, 2 GB RAM, no guest additions.

Kernel errors examples:

kernel BUG at mm/vmscan.c:1083!
invalid opcode: 0000 [#1] SMP
BUG: soft lockup - CPU#0 stuck for 67s! [updater:31994]

VirtualBox and console logs attached. Thanks for suggestions.

[mpack]

You have lots of this in the log :-

47:46:17.299835 TM: Giving up catch-up attempt at a 60 004 717 530 ns lag; new total: 12 538 020 201 981 ns
47:57:28.689517 TM: Giving up catch-up attempt at a 60 001 553 368 ns lag; new total: 12 598 021 755 349 ns
48:02:20.437494 TM: Giving up catch-up attempt at a 60 000 407 586 ns lag; new total: 12 658 022 162 935 ns
48:13:18.222674 TM: Giving up catch-up attempt at a 60 004 200 350 ns lag; new total: 12 718 026 363 285 ns
48:22:51.722697 TM: Giving up catch-up attempt at a 60 003 556 963 ns lag; new total: 12 778 029 920 248 ns
48:26:55.225640 TM: Giving up catch-up attempt at a 60 001 736 593 ns lag; new total: 12 838 031 656 841 ns
48:39:04.361371 TM: Giving up catch-up attempt at a 60 000 463 181 ns lag; new total: 12 898 032 120 022 ns
48:52:18.813530 TM: Giving up catch-up attempt at a 60 002 636 844 ns lag; new total: 12 958 034 756 866 ns
48:57:32.937526 TM: Giving up catch-up attempt at a 60 000 090 163 ns lag; new total: 13 018 034 847 029 ns
49:06:55.400391 TM: Giving up catch-up attempt at a 60 000 264 146 ns lag; new total: 13 078 035 111 175 ns

I suspect something outside the VM may be eating up a lot of CPU (like another VM), or maybe a backup process - but you should probably raise a bugtracker ticket for this.