With a large number of servers, it becomes difficult to fill even the nodes section.
I suggest to add command to ccg, that allows to generate nodes section for cluster.yaml or add new nodes to existed cluster.yaml by patterns. Patten should support range syntax.
Example: 192.168.0.[10-18]:20000/opt/disk[1-12]
With this pattern nodes section should have 8 nodes (192.168.0.10:20000, 192.168.0.11:20000, ..., 192.168.0.18:20000) each of which should have 12 disks (/opt/disk1, /opt/disk2, ..., /opt/disk12).
Node name should be generated from node address by replacing non digit or letters symbols to _. Example: 192_168_0_10.
Disk name should be generated with auto increment on every node. Example: disk1, disk2, ...
Another valid pattern examples: node-[1-8]:20000/opt/l[1-2]/d[1-4], 192.168.[0-4].[10-12]:20000/disk1
With a large number of servers, it becomes difficult to fill even the
nodessection.I suggest to add command to ccg, that allows to generate
nodessection forcluster.yamlor add new nodes to existedcluster.yamlby patterns. Patten should support range syntax.Example:
192.168.0.[10-18]:20000/opt/disk[1-12]With this pattern
nodessection should have 8 nodes (192.168.0.10:20000,192.168.0.11:20000, ...,192.168.0.18:20000) each of which should have 12 disks (/opt/disk1,/opt/disk2, ...,/opt/disk12).Node name should be generated from node address by replacing non digit or letters symbols to
_. Example:192_168_0_10.Disk name should be generated with auto increment on every node. Example:
disk1,disk2, ...Another valid pattern examples:
node-[1-8]:20000/opt/l[1-2]/d[1-4],192.168.[0-4].[10-12]:20000/disk1