[PVE-User] Creating Ceph Cluster

[Markus Dellermann]

Am Freitag, 26. Februar 2016, 23:28:17 CET schrieb Vadim Bulst:
> Dear List,
> 
> after 2 years running a PVE-Cluster of four nodes with Ceph-RBD it got
> time to for PVE major release Four. We added a fifth node with the same
> version of PVE (3.4) and local ZFS-Volume and migrated all important VMs
> to this node and ZFS-Storage. After this was done we removed all old
> nodes from the existing cluster.
> 
> Now we made a dist-upgrade to Debian Jessie and PVE 4.1 on the fifth
> (new) node. The old node we installed completely new to PVE 4.1 and
> joined to the updated fifth node.
> 
> Everything fine so far. Installation and upgrade went pretty straight
> forward.
> 
> PVE-Cluster is up and running again.
> 
> Now we where going to install the CEPH-components. Added the Hammer-repo
> of ceph.com , installed all in the wiki mentioned packages.
> 
> pveceph init --network 172.18.117.0/24 was successfully executed.
> 
> Now we tried to add the first cephmon on one of the new installed nodes
> with no luck. We executed the following command:
> 
> pveceph createmon
> 
> The error message is:
> 
> unable to find local address within network '172.18.117.0/24'
> 
> The only node we were able to add is the from PVE 3.4 upgraded one.
> 
> This is our network configuration:
> 
> # Loopback interface
> auto lo
> iface lo inet loopback
> 
> # Bond eth0 and eth1 together
> allow-vmbr1 bond0
> iface bond0 inet manual
>          ovs_bridge vmbr1
>          ovs_type OVSBond
>          ovs_bonds eth1 eth0
>          pre-up ( ifconfig eth0 mtu 1500 && ifconfig eth1 mtu 1500 )
>          ovs_options bond_mode=balance-slb
>          mtu 1500
> 
> # Bridge for our bond and vlan virtual interfaces (our VMs will
> # also attach to this bridge)
> auto vmbr1
> allow-ovs vmbr1
> iface vmbr1 inet manual
>          ovs_type OVSBridge
>          ovs_ports bond0 vlan888
>          mtu 1500
> 
> allow-vmbr1 vlan888
> iface vlan888 inet static
>          ovs_type OVSIntPort
>          ovs_bridge vmbr1
>          ovs_options tag=888
>          ovs_extra set interface ${IFACE}
> external-ids:iface-id=$(hostname -s)-${IFACE}-vif
>          address 172.18.117.93
>          netmask 255.255.255.0
>          gateway 172.18.117.254
>          mtu 1500
> 
> 
> What went wrong? Does anybody has a supposition?
> 
> Cheers,
> 
> Vadim
> 
Hi Vadim,
i think, you should look at the output from 
"systemctl status networking.service -l "

Maybe there is a problem in your /etc/network/interfaces...

Markus
> _______________________________________________
> pve-user mailing list
> pve-user at pve.proxmox.com
> http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-user