Heads up: New Ceph images require x86-64-v2 and possibly a qemu config change for virtual servers
Hey all, The upcoming community Ceph container images will be based on CentOS 9. In our Clyso CI testing lab we learned that el9-based images won't run on some (default) qemu VMs. Where our el8-based images run well, our new el9-based images get: error during bootstrap: Fatal glibc error: CPU does not support x86-64-v2 This is because RHEL 9 (and clones) added a hard requirement on x86-64-v2: * https://srobb.net/rhel9.html * https://forums.rockylinux.org/t/el9-will-require-x86-64-v2-support/5311 * https://developers.redhat.com/blog/2021/01/05/building-red-hat-enterprise-li... * https://access.redhat.com/solutions/6833751 The reason for the failure in our testing is that some of our virtual nodes use qemu's default CPU emulation (qemu64), which does not support x86-64-v2. The solution for these nodes is to switch to "host" or something else listed at https://www.qemu.org/docs/master/system/i386/cpu.html I would encourage cephadm/Rook users with daemons on VM hosts to verify their CPU modes on their VMs prior to any upgrades, and make the necessary changes to the hypervisor configuration ahead of time. Best, Dan -- Dan van der Ster CTO Clyso GmbH p: +49 89 215252722 | a: Vancouver, Canada w: https://clyso.com | e: dan.vanderster@clyso.com
Thanks a lot for the heads up, Dan! Zitat von Dan van der Ster <dan.vanderster@clyso.com>:
Hey all,
The upcoming community Ceph container images will be based on CentOS 9.
In our Clyso CI testing lab we learned that el9-based images won't run on some (default) qemu VMs. Where our el8-based images run well, our new el9-based images get:
error during bootstrap: Fatal glibc error: CPU does not support x86-64-v2
This is because RHEL 9 (and clones) added a hard requirement on x86-64-v2:
* https://srobb.net/rhel9.html * https://forums.rockylinux.org/t/el9-will-require-x86-64-v2-support/5311 * https://developers.redhat.com/blog/2021/01/05/building-red-hat-enterprise-li... * https://access.redhat.com/solutions/6833751
The reason for the failure in our testing is that some of our virtual nodes use qemu's default CPU emulation (qemu64), which does not support x86-64-v2. The solution for these nodes is to switch to "host" or something else listed at https://www.qemu.org/docs/master/system/i386/cpu.html
I would encourage cephadm/Rook users with daemons on VM hosts to verify their CPU modes on their VMs prior to any upgrades, and make the necessary changes to the hypervisor configuration ahead of time.
Best, Dan
-- Dan van der Ster CTO
Clyso GmbH p: +49 89 215252722 | a: Vancouver, Canada w: https://clyso.com | e: dan.vanderster@clyso.com _______________________________________________ ceph-users mailing list -- ceph-users@ceph.io To unsubscribe send an email to ceph-users-leave@ceph.io
+1 to this, also ran into this in our lab testing. Thanks for sharing this information! Regards, Bailey
-----Original Message----- From: Eugen Block <eblock@nde.ag> Sent: July 18, 2024 3:55 AM To: ceph-users@ceph.io Subject: [ceph-users] Re: Heads up: New Ceph images require x86-64-v2 and possibly a qemu config change for virtual servers
Thanks a lot for the heads up, Dan!
Zitat von Dan van der Ster <dan.vanderster@clyso.com>:
Hey all,
The upcoming community Ceph container images will be based on CentOS
In our Clyso CI testing lab we learned that el9-based images won't run on some (default) qemu VMs. Where our el8-based images run well, our new el9-based images get:
error during bootstrap: Fatal glibc error: CPU does not support x86-64-v2
This is because RHEL 9 (and clones) added a hard requirement on x86-64-
v2:
* https://srobb.net/rhel9.html * https://forums.rockylinux.org/t/el9-will-require-x86-64-v2-support/531 1 * https://developers.redhat.com/blog/2021/01/05/building-red-hat-enterpr ise-linux-9-for-the-x86-64-v2-microarchitecture-level * https://access.redhat.com/solutions/6833751
The reason for the failure in our testing is that some of our virtual nodes use qemu's default CPU emulation (qemu64), which does not support x86-64-v2. The solution for these nodes is to switch to "host" or something else listed at https://www.qemu.org/docs/master/system/i386/cpu.html
I would encourage cephadm/Rook users with daemons on VM hosts to verify their CPU modes on their VMs prior to any upgrades, and make the necessary changes to the hypervisor configuration ahead of time.
Best, Dan
-- Dan van der Ster CTO
Clyso GmbH p: +49 89 215252722 | a: Vancouver, Canada w: https://clyso.com | e: dan.vanderster@clyso.com _______________________________________________ ceph-users mailing list -- ceph-users@ceph.io To unsubscribe send an email to ceph-users-leave@ceph.io
_______________________________________________ ceph-users mailing list -- ceph-users@ceph.io To unsubscribe send an email to ceph-users-leave@ceph.io
participants (3)
-
Bailey Allison
-
Dan van der Ster
-
Eugen Block