Not able to install Ceph on Rocky Linux 9.6

Main error is nothing provides libcrypto.so.3(OPENSSL_3.4.0)(64bit) needed by ceph-common-2:19.2.3-0.el9.x86_64 from ceph_stable_x86_64

# dnf install ceph-common
Last metadata expiration check: 0:13:50 ago on Fri 22 Aug 2025 02:32:28 PM.
Error: 
 Problem: cannot install the best candidate for the job
  - nothing provides libcrypto.so.3(OPENSSL_3.4.0)(64bit) needed by ceph-common-2:19.2.3-0.el9.x86_64 from ceph_stable_x86_64
(try to add '--skip-broken' to skip uninstallable packages or '--nobest' to use not only best candidate packages)

I have opened an issue here, but curious to know if anyone here has any input: Not able to install on Rocky Linux 9.6 `nothing provides libcrypto.so.3(OPENSSL_3.4.0)(64bit) needed by ceph-common-2:19.2.3-0.el9.x86_64 from ceph_stable_x86_64` · Issue #345 · ceph/cephadm-ansible · GitHub

Host is fully up to date and brand new

# cat /etc/os-release 
NAME="Rocky Linux"
VERSION="9.6 (Blue Onyx)"
ID="rocky"
ID_LIKE="rhel centos fedora"
VERSION_ID="9.6"
PLATFORM_ID="platform:el9"
PRETTY_NAME="Rocky Linux 9.6 (Blue Onyx)"
ANSI_COLOR="0;32"
LOGO="fedora-logo-icon"
CPE_NAME="cpe:/o:rocky:rocky:9::baseos"
HOME_URL="https://rockylinux.org/"
VENDOR_NAME="RESF"
VENDOR_URL="https://resf.org/"
BUG_REPORT_URL="https://bugs.rockylinux.org/"
SUPPORT_END="2032-05-31"
ROCKY_SUPPORT_PRODUCT="Rocky-Linux-9"
ROCKY_SUPPORT_PRODUCT_VERSION="9.6"
REDHAT_SUPPORT_PRODUCT="Rocky Linux"
REDHAT_SUPPORT_PRODUCT_VERSION="9.6"
# dnf upgrade
Last metadata expiration check: 0:14:56 ago on Fri 22 Aug 2025 04:54:56 PM.
Dependencies resolved.
Nothing to do.
Complete!

I’ve run into the same dependency errors on an existing ceph installation sometime in April or May (I think – it was before 9.6 was released). The reason for the error is that packages in ceph’s repositories are built against CentOS Stream which (in this case) is a little more than just ahead of proper enterprise linux releases.

I think I saw somewhere that the required openssl should be in RHEL/Rocky 9.7.

I don’t have the link handy, but a while back I found some reference to this issue on a ceph forum or mailing list. My impression was that there was no real willingness to address this as a worthwhile problem – the general attitude was that everyone should just use container-based deployments.

But does this not prevent the use of any ceph comands to do any sort of configuring on the OSD for example?

By manually installing podman I am able to add the host to the cluster using ceph gui, but I am not able to run any ceph commands on that host to configure anything.

I believe cephadmis one of the container-centered tools that the ceph team seems to be interested in supporting an promoting. So hopefully your bug report should get some traction and not just a recommendation to run cephadm itself as a container.

This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.