CentOS Question posted in
The official documentation can be found here.

ceph-deploy install –release octopus on centos7 tries to load non existing ceph-release-1-0 and ceph-deploy doesn't like centos8

I’m in a dead lock! Would like to extend a ceph nautilus cluster running on centos7 by a centos8 node.
I used ceph-deploy for deployment. Now the problems:

  • ceph-deploy (2.0.1) does not recognize centos8 (8.3.2011) as a supported plattform:

    [ceph_deploy][ERROR ] UnsupportedPlatform: Platform is not supported: CentOS Linux 8.3.2011

  • I cannot switch to cephadm since it does not support nautilus

  • I cannot upgrade to octopus on centos7, since ceph-deploy tries to
    load a non existing ceph-release-1-0:

    [c1][WARNIN] Cannot open: https://download.ceph.com/rpm-octopus/el7/noarch/ceph-release-1-0.el7.noarch.rpm. Skipping.

So there is no way out! Any help is appreciated.

Magnus

Tags:

2

Answers


  1. Chosen as BEST ANSWER
    0

    Fixed ceph-deploy the hard way!

    cd /usr/lib/python2.7/site-packages/ceph_deploy/hosts/centos

sed -i -e "s/.*ceph-release-1-0.*/'{url}noarch/ceph-release-1-1.el7.noarch.rpm'.format(url=url),/" install.py

mv install.pyo install.pyox
mv install.pyc install.pycx

    Magnus


  2. 0

    Magnus, I am in the same boat a you. I am using Nautilus, but want to upgrade and ran into the same error. In a bit of catch-22 with CentOS7 and Nautilus vs CentOS8 and Octopus. I setup a Virtual Machine test cluster and upgraded Nautilus to Octopus on CentOS7. My understanding is the dashboard will not work correctly but the cluster will continue to function. My plan is to upgrade my production CentOS7 cluster to Octopus, then add CentOS8 nodes.
    While I was able to get Octopus on CentOS7, I cannot for the life of me get cephadm running correctly. cephadm added the host but getting error No daemons reported. I don’t know enough about containers. If I cannot perform this upgrade with confidence I am afraid I will have to build a new cluster and migrate that way.
    Anyway hope this helps. Run tests first.

    **On Each Node:

        sudo yum install python3 -y
    sudo yum install podman -y
    sudo yum install dnf -y

    **Change version name ex:Nautilus to Octopus

        sudo vi /etc/yum.repos.d/ceph.repo

    **Check Version – should be Nautilus

        Ceph -v

    sudo dnf update -y

    **Check Version – should now be Octopus

        Ceph -v

    **restart services

        sudo systemctl restart ceph-mon.target
    sudo systemctl restart ceph-mgr.target
    sudo systemctl restart ceph-osd.target

    ceph tell mon.* version
    ceph tell osd.* version
    ceph mon stat

    **Once all monitors are up, verify that the monitor upgrade is complete by looking for the octopus string in the mon map. The command:

        ceph mon dump | grep min_mon_release
    Login or Signup to reply.
Please signup or login to give your own answer.

Back To Top
Search