On 2022-11-08 15:49, Orion Poplawski wrote:
On 11/8/22 13:12, Simon Matter wrote: Is anyone else experiencing trouble with kernel-3.10.0-1160.80.1.el7.x86_64? I'm seeing a kernel panics in the kvm module on one of our VM hosts with it. I did notice a new libvirt update as well, but it seems to work fine with the older kernel (.76.1). Where did you get the .80.1 kernel from? I'm a bit confused because I can only see .76.1 on my systems. Simon I'm actually running Scientific Linux, which seems to be a little ahead here.Probably not related, but vmlinuz-4.18.0-372.32.1.el8_6.x86_64 (AlmaLinux 8.6) had a kernel panic on a Intel Xeon E5504 processor, but works fine on Xeon E56XX processors (5620 specifically in our tests). * Believe from kernel version, the original email is for Centos 7, possible the same change that went into the EL8 kernel went also to EL7 Triggered right after I used virsh start to start the vm. (Caused Black Screen, and reboot. Found this in the crashed kernel logs on /var/crash) Works fine on vmlinuz-4.18.0-372.26.1.el8_6.x86_64 and vmlinuz-4.18.0-372.16.1.el8_6.x86_64
In fact, related - on our systems attempt to start qemu-kvm VM under kernel-3.10.0-1160.80.1.el7.x86_64 freezes any of 5 workstations with dual E5507 (all worked OK under kernel-3.10.0-1160.76.1.el7.x86_64 and any previous version). The workstations with E5-2609, E5-2650 or E5-2630 are not affected - all of them run qemu-kvm VM under kernel-3.10.0-1160.80.1.el7.x86_64 without problems.
Any ideas?
Regards,
Petko
On 11/15/22 7:50 PM, Petko Alov wrote:
On 2022-11-08 15:49, Orion Poplawski wrote:
On 11/8/22 13:12, Simon Matter wrote:
Is anyone else experiencing trouble with kernel-3.10.0-1160.80.1.el7.x86_64?
I'm seeing a kernel panics in the kvm module on one of our VM hosts with
it.
I did notice a new libvirt update as well, but it seems to work fine with
the older kernel (.76.1).
Where did you get the .80.1 kernel from? I'm a bit confused because I can
only see .76.1 on my systems.
Simon
I'm actually running Scientific Linux, which seems to be a little ahead here. Probably not related, but vmlinuz-4.18.0-372.32.1.el8_6.x86_64 (AlmaLinux 8.6) had a kernel panic on a Intel Xeon E5504 processor, but works fine on Xeon E56XX processors (5620 specifically in our tests). * Believe from kernel version, the original email is for Centos 7, possible the same change that went into the EL8 kernel went also to EL7 Triggered right after I used virsh start to start the vm. (Caused Black Screen, and reboot. Found this in the crashed kernel logs on /var/crash) Works fine on vmlinuz-4.18.0-372.26.1.el8_6.x86_64 and vmlinuz-4.18.0-372.16.1.el8_6.x86_64
In fact, related - on our systems attempt to start qemu-kvm VM under kernel-3.10.0-1160.80.1.el7.x86_64 freezes any of 5 workstations with dual E5507 (all worked OK under kernel-3.10.0-1160.76.1.el7.x86_64 and any previous version). The workstations with E5-2609, E5-2650 or E5-2630 are not affected - all of them run qemu-kvm VM under kernel-3.10.0-1160.80.1.el7.x86_64 without problems.
Thankfully, no problems with 2x Xeon E5-2667 v2 and CentOS 7.9 kernel 3.10.0-1160.80.1.el7.x86_64.
On the other hand --- I just upgraded one of my CentOS7 systems to 3.10.0-1160.80.1 today. With that kernel it will not run any KVM guest. The system hangs - HARD - within one or two seconds of issuing the "virsh start ..." command. It hangs so hard that a power-off restart using the back panel switch is required. It does not log any error I can find, nor display any message about a kernel panic.
Rebooting with 3.10.0-1160.76.1 works fine.
The system is an older SuperMicro mainboard with a Q9400 processor.
=============== Bill Gee
On 11/16/22 12:38, Chris Schanzle via CentOS wrote:
On 11/15/22 7:50 PM, Petko Alov wrote:
On 2022-11-08 15:49, Orion Poplawski wrote:
On 11/8/22 13:12, Simon Matter wrote:
Is anyone else experiencing trouble with kernel-3.10.0-1160.80.1.el7.x86_64?
I'm seeing a kernel panics in the kvm module on one of our VM hosts with
it.
I did notice a new libvirt update as well, but it seems to work fine with
the older kernel (.76.1).
Where did you get the .80.1 kernel from? I'm a bit confused because I can
only see .76.1 on my systems.
Simon
I'm actually running Scientific Linux, which seems to be a little ahead here. Probably not related, but vmlinuz-4.18.0-372.32.1.el8_6.x86_64 (AlmaLinux 8.6) had a kernel panic on a Intel Xeon E5504 processor, but works fine on Xeon E56XX processors (5620 specifically in our tests). * Believe from kernel version, the original email is for Centos 7, possible the same change that went into the EL8 kernel went also to EL7 Triggered right after I used virsh start to start the vm. (Caused Black Screen, and reboot. Found this in the crashed kernel logs on /var/crash) Works fine on vmlinuz-4.18.0-372.26.1.el8_6.x86_64 and vmlinuz-4.18.0-372.16.1.el8_6.x86_64
In fact, related - on our systems attempt to start qemu-kvm VM under kernel-3.10.0-1160.80.1.el7.x86_64 freezes any of 5 workstations with dual E5507 (all worked OK under kernel-3.10.0-1160.76.1.el7.x86_64 and any previous version). The workstations with E5-2609, E5-2650 or E5-2630 are not affected - all of them run qemu-kvm VM under kernel-3.10.0-1160.80.1.el7.x86_64 without problems.
Thankfully, no problems with 2x Xeon E5-2667 v2 and CentOS 7.9 kernel 3.10.0-1160.80.1.el7.x86_64.
CentOS mailing list CentOS@centos.org https://lists.centos.org/mailman/listinfo/centos
Has this been solved? Is there a bug report I can follow? Thank you for letting me know,
Valère Binet
On Wed, Nov 16, 2022 at 2:23 PM Bill Gee bgee@campercaver.net wrote:
On the other hand --- I just upgraded one of my CentOS7 systems to 3.10.0-1160.80.1 today. With that kernel it will not run any KVM guest. The system hangs - HARD - within one or two seconds of issuing the "virsh start ..." command. It hangs so hard that a power-off restart using the back panel switch is required. It does not log any error I can find, nor display any message about a kernel panic.
Rebooting with 3.10.0-1160.76.1 works fine.
The system is an older SuperMicro mainboard with a Q9400 processor.
=============== Bill Gee
On 11/16/22 12:38, Chris Schanzle via CentOS wrote:
On 11/15/22 7:50 PM, Petko Alov wrote:
On 2022-11-08 15:49, Orion Poplawski wrote:
On 11/8/22 13:12, Simon Matter wrote: Is anyone else experiencing trouble with kernel-3.10.0-1160.80.1.el7.x86_64? I'm seeing a kernel panics in the kvm module on one ofour VM hosts with
it. I did notice a new libvirt update as well, but it seemsto work fine with
the older kernel (.76.1). Where did you get the .80.1 kernel from? I'm a bit confusedbecause I can
only see .76.1 on my systems. Simon I'm actually running Scientific Linux, which seems to be a littleahead here. Probably not related, but vmlinuz-4.18.0-372.32.1.el8_6.x86_64 (AlmaLinux 8.6) had a kernel panic on a Intel Xeon E5504 processor, but works fine on Xeon E56XX processors (5620 specifically in our tests). * Believe from kernel version, the original email is for Centos 7, possible the same change that went into the EL8 kernel went also to EL7 Triggered right after I used virsh start to start the vm. (Caused Black Screen, and reboot. Found this in the crashed kernel logs on /var/crash) Works fine on vmlinuz-4.18.0-372.26.1.el8_6.x86_64 and vmlinuz-4.18.0-372.16.1.el8_6.x86_64
In fact, related - on our systems attempt to start qemu-kvm VM under kernel-3.10.0-1160.80.1.el7.x86_64 freezes any of 5 workstations with dual E5507 (all worked OK under kernel-3.10.0-1160.76.1.el7.x86_64 and any previous version). The workstations with E5-2609, E5-2650 or E5-2630 are not affected - all of them run qemu-kvm VM under kernel-3.10.0-1160.80.1.el7.x86_64 without problems.
Thankfully, no problems with 2x Xeon E5-2667 v2 and CentOS 7.9 kernel 3.10.0-1160.80.1.el7.x86_64.
CentOS mailing list CentOS@centos.org https://lists.centos.org/mailman/listinfo/centos
CentOS mailing list CentOS@centos.org https://lists.centos.org/mailman/listinfo/centos
The issue was filed at bugzilla@redhat.com, no results yet...
[Bug 2143438] kernel 3.10.0-1160.80.1.el7.x86_64 on Xeon E55xx crashes upon KVM startup
https://bugzilla.redhat.com/show_bug.cgi?id=2143438
Regards,
Petko
On 12/14/22 4:47 PM, Valere Binet wrote:
Has this been solved? Is there a bug report I can follow? Thank you for letting me know,
Valère Binet
On Wed, Nov 16, 2022 at 2:23 PM Bill Gee bgee@campercaver.net wrote:
On the other hand -Subscribe to CentOS by filling out the following form. You will be sent email requesting confirmation, to prevent others from gratuitously subscribing you. This is a hidden list, which means that the list of members is available only to the list administrator. -- I just upgraded one of my CentOS7 systems to 3.10.0-1160.80.1 today. With that kernel it will not run any KVM guest. The system hangs - HARD - within one or two seconds of issuing the "virsh start ..." command. It hangs so hard that a power-off restart using the back panel switch is required. It does not log any error I can find, nor display any message about a kernel panic.
Rebooting with 3.10.0-1160.76.1 works fine.
The system is an older SuperMicro mainboard with a Q9400 processor.
=============== Bill Gee
On 11/16/22 12:38, Chris Schanzle via CentOS wrote:
On 11/15/22 7:50 PM, Petko Alov wrote:
On 2022-11-08 15:49, Orion Poplawski wrote:
On 11/8/22 13:12, Simon Matter wrote: Is anyone else experiencing trouble with kernel-3.10.0-1160.80.1.el7.x86_64? I'm seeing a kernel panics in the kvm module on one ofour VM hosts with
it. I did notice a new libvirt update as well, but it seemsto work fine with
the older kernel (.76.1). Where did you get the .80.1 kernel from? I'm a bit confusedbecause I can
only see .76.1 on my systems. Simon I'm actually running Scientific Linux, which seems to be a littleahead here. Probably not related, but vmlinuz-4.18.0-372.32.1.el8_6.x86_64 (AlmaLinux 8.6) had a kernel panic on a Intel Xeon E5504 processor, but works fine on Xeon E56XX processors (5620 specifically in our tests). * Believe from kernel version, the original email is for Centos 7, possible the same change that went into the EL8 kernel went also to EL7 Triggered right after I used virsh start to start the vm. (Caused Black Screen, and reboot. Found this in the crashed kernel logs on /var/crash) Works fine on vmlinuz-4.18.0-372.26.1.el8_6.x86_64 and vmlinuz-4.18.0-372.16.1.el8_6.x86_64
In fact, related - on our systems attempt to start qemu-kvm VM under kernel-3.10.0-1160.80.1.el7.x86_64 freezes any of 5 workstations with dual E5507 (all worked OK under kernel-3.10.0-1160.76.1.el7.x86_64 and any previous version). The workstations with E5-2609, E5-2650 or E5-2630 are not affected - all of them run qemu-kvm VM under kernel-3.10.0-1160.80.1.el7.x86_64 without problems.
Thankfully, no problems with 2x Xeon E5-2667 v2 and CentOS 7.9 kernel 3.10.0-1160.80.1.el7.x86_64.
CentOS mailing list CentOS@centos.org https://lists.centos.org/mailman/listinfo/centos
CentOS mailing list CentOS@centos.org https://lists.centos.org/mailman/listinfo/centos
CentOS mailing list CentOS@centos.org https://lists.centos.org/mailman/listinfo/centos
Finally, the KVM hanging issue on E55xx CPUs is resolved with kernel.x86_64-3.10.0-1160.88.1.el7!
On 11/16/22 02:50, Petko Alov wrote:
On 2022-11-08 15:49, Orion Poplawski wrote:
On 11/8/22 13:12, Simon Matter wrote:
Is anyone else experiencing trouble with kernel-3.10.0-1160.80.1.el7.x86_64?
I'm seeing a kernel panics in the kvm module on one of our VM hosts with
it.
I did notice a new libvirt update as well, but it seems to work fine with
the older kernel (.76.1).
Where did you get the .80.1 kernel from? I'm a bit confused because I can
only see .76.1 on my systems.
Simon
I'm actually running Scientific Linux, which seems to be a little ahead here. Probably not related, but vmlinuz-4.18.0-372.32.1.el8_6.x86_64 (AlmaLinux 8.6) had a kernel panic on a Intel Xeon E5504 processor, but works fine on Xeon E56XX processors (5620 specifically in our tests). * Believe from kernel version, the original email is for Centos 7, possible the same change that went into the EL8 kernel went also to EL7 Triggered right after I used virsh start to start the vm. (Caused Black Screen, and reboot. Found this in the crashed kernel logs on /var/crash) Works fine on vmlinuz-4.18.0-372.26.1.el8_6.x86_64 and vmlinuz-4.18.0-372.16.1.el8_6.x86_64
In fact, related - on our systems attempt to start qemu-kvm VM under kernel-3.10.0-1160.80.1.el7.x86_64 freezes any of 5 workstations with dual E5507 (all worked OK under kernel-3.10.0-1160.76.1.el7.x86_64 and any previous version). The workstations with E5-2609, E5-2650 or E5-2630 are not affected - all of them run qemu-kvm VM under kernel-3.10.0-1160.80.1.el7.x86_64 without problems.
Any ideas?
Regards,
Petko
-
Bill Gee -
Chris Schanzle -
Petko Alov -
Valere Binet