Hi all, I'm having trouble bootstrapping new cephadm ceph cluster. Background 1. I'm running withing an isolated environment, so I'm running my own registry (with ceph:v17.2.2) and my own package repository. 2. I'm also running it on a proxmox debian installation. 3. I had no problem bootstrapping a cluster on pacific, now I'm trying to start a new cluster on quincy. 4. cephadm has been both copied to /usr/bin and chmodded as well as installed using apt install The issue is that cephadm fails when trying to add the initial cephadm host during the bootstrap process. I'm using this cephadm command: cephadm --image <local registry>/ceph/ceph:v17 bootstrap --mon-ip <local ip> --cluster-network <cluster subnet/24> I get a "non-zero exit code 22" from podman and ends: ERROR EINVAL: Unable to reach remote host <hostname>. Process exited with non-zero exit status 127. The hostname and fqdn resolves locally and in dns to the ip address. I can ping the ip, hostname and fqdn from the command line. Running ssh <hostname> or ssh <ip> in a cephadm shell resolves (<hostname> goes to 127.0.0.1) and works (there is no ping, dig or nslookup in the container). If I spin up a VM with a debian cloud image and install cephadm via apt, I can bootstrap a new cluster. I still can't add the original host to the cluster, the same error (unable to reach host <hostname>). Ideas? Issue with 17.2.2? Thanks! _______________________________________________ ceph-users mailing list -- ceph-users@xxxxxxx To unsubscribe send an email to ceph-users-leave@xxxxxxx
 |