KVM Guest Lock up (100%) again!

Hello all,

Another lock up again this evening :( am wondering whether should consider upgrading the kernel to 3.7.10 and the latest version of KVM. Thoughts ?

Thanks.

----- Original Message -----
To: [email protected]
Sent: Thursday, 4 April, 2013 3:36:11 PM
Subject: KVM Guest Lock up (100%) again!

One of my KVM guests locked up again at 100% CPU! Any thoughts on how I can diagnose it ? We would love to put into production but am very concerned about the current stability. I have tried to re-direct the console, through screen, to see whether there is a spin lock that is causing the problem; though all I got in the log file was a login prompt. What is the correct way of redirecting the console in KVM on a CentOS 6.4 system please ? Thanks. --
To unsubscribe from this list: send the line "unsubscribe kvm" in
the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
--
To unsubscribe from this list: send the line "unsubscribe kvm" in
the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

Hello all,

Another lock up again this evening :( am wondering whether should consider upgrading the kernel to 3.7.10 and the latest version of KVM. Thoughts ?

Thanks.

----- Original Message -----
To: [email protected]
Sent: Thursday, 4 April, 2013 3:36:11 PM
Subject: KVM Guest Lock up (100%) again!

One of my KVM guests locked up again at 100% CPU! Any thoughts on how I can diagnose it ? We would love to put into production but am very concerned about the current stability. I have tried to re-direct the console, through screen, to see whether there is a spin lock that is causing the problem; though all I got in the log file was a login prompt. What is the correct way of redirecting the console in KVM on a CentOS 6.4 system please ? Thanks. --
To unsubscribe from this list: send the line "unsubscribe kvm" in
the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
--
To unsubscribe from this list: send the line "unsubscribe kvm" in
the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

Well this is still happening ... I have tried to isolate what could be causing but not much luck yet. Thought the VMs may have been IO bound but that not the case and even tried upping the vCPU allocation from one to two as plenty of head room. When it locks up I see this on a strace:

[pid  1343] read(14, 0x7fff82aeb360, 4096) = -1 EAGAIN (Resource temporarily unavailable)
[pid  1343] read(7, "\0", 512)          = 1
[pid  1343] read(7, 0x7fff82aec160, 512) = -1 EAGAIN (Resource temporarily unavailable)
[pid  1343] select(26, [7 10 13 14 16 17 22 25], [], [], {1, 0}) = 1 (in [16], left {0, 999981})
[pid  1343] read(16, "\16\0\0\0\0\0\0\0\376\377\377\377\0\0\0\0\0\0\0\0\0\0\0\0\1\0\0\0\0\0\0\0"..., 128) = 128
[pid  1343] rt_sigaction(SIGALRM, NULL, {0x7f210b2c0510, ~[KILL STOP RTMIN RT_1], SA_RESTORER, 0x7f210ac22500}, 8) = 0
[pid  1343] write(8, "\0", 1)           = 1
[pid  1343] write(15, "\1\0\0\0\0\0\0\0", 8) = 8
[pid  1343] read(16, 0x7fff82aec2d0, 128) = -1 EAGAIN (Resource temporarily unavailable)
[pid  1343] timer_gettime(0x1, {it_interval={0, 0}, it_value={0, 0}}) = 0
[pid  1343] timer_settime(0x1, 0, {it_interval={0, 0}, it_value={0, 656000000}}, NULL) = 0
[pid  1343] select(26, [7 10 13 14 16 17 22 25], [], [], {1, 0}) = 2 (in [7 14], left {0, 999998})
[pid  1343] read(14, "\1\0\0\0\0\0\0\0", 4096) = 8
[pid  1343] read(14, 0x7fff82aeb360, 4096) = -1 EAGAIN (Resource temporarily unavailable)
[pid  1343] read(7, "\0", 512)          = 1
[pid  1343] read(7, 0x7fff82aec160, 512) = -1 EAGAIN (Resource temporarily unavailable)

Does that shed any light ? Trying to find a how to for upgrading to the latest KVM/QEMU.

Well this is still happening ... I have tried to isolate what could be causing but not much luck yet. Thought the VMs may have been IO bound but that not the case and even tried upping the vCPU allocation from one to two as plenty of head room. When it locks up I see this on a strace:

[pid  1343] read(14, 0x7fff82aeb360, 4096) = -1 EAGAIN (Resource temporarily unavailable)
[pid  1343] read(7, "\0", 512)          = 1
[pid  1343] read(7, 0x7fff82aec160, 512) = -1 EAGAIN (Resource temporarily unavailable)
[pid  1343] select(26, [7 10 13 14 16 17 22 25], [], [], {1, 0}) = 1 (in [16], left {0, 999981})
[pid  1343] read(16, "\16\0\0\0\0\0\0\0\376\377\377\377\0\0\0\0\0\0\0\0\0\0\0\0\1\0\0\0\0\0\0\0"..., 128) = 128
[pid  1343] rt_sigaction(SIGALRM, NULL, {0x7f210b2c0510, ~[KILL STOP RTMIN RT_1], SA_RESTORER, 0x7f210ac22500}, 8) = 0
[pid  1343] write(8, "\0", 1)           = 1
[pid  1343] write(15, "\1\0\0\0\0\0\0\0", 8) = 8
[pid  1343] read(16, 0x7fff82aec2d0, 128) = -1 EAGAIN (Resource temporarily unavailable)
[pid  1343] timer_gettime(0x1, {it_interval={0, 0}, it_value={0, 0}}) = 0
[pid  1343] timer_settime(0x1, 0, {it_interval={0, 0}, it_value={0, 656000000}}, NULL) = 0
[pid  1343] select(26, [7 10 13 14 16 17 22 25], [], [], {1, 0}) = 2 (in [7 14], left {0, 999998})
[pid  1343] read(14, "\1\0\0\0\0\0\0\0", 4096) = 8
[pid  1343] read(14, 0x7fff82aeb360, 4096) = -1 EAGAIN (Resource temporarily unavailable)
[pid  1343] read(7, "\0", 512)          = 1
[pid  1343] read(7, 0x7fff82aec160, 512) = -1 EAGAIN (Resource temporarily unavailable)

Does that shed any light ? Trying to find a how to for upgrading to the latest KVM/QEMU.