On Wednesday, December 10, 2025 3:07:34 AM Eastern Standard Time Robert Sander via ceph-users wrote: > Am 09.12.25 um 8:05 PM schrieb John Mulligan: > ´ > > > I have filed a PR [1]. I have tried to mark all options that can be > > supplied more than once as such in each items description list. > > > Thank you. How is the interface determined where a public address has to > be bound to when the cli argument looks like this? > > --public-address=192.168.76.110/24 >
This simple form is designed to work well when the cluster is configured for a "common subnet" scenario. Let's imagine our ceph cluster is 3-nodes large and we'll be deploying an smb instance for each one. For example: node A - 192.0.2.100/24 (eth0) node B - 192.0.2.101/24 (eth0) node C - 192.0.2.102/24 (eth0) We'll use public address 192.0.2.110/24. Let's assume that the address is to be assigned to node A. The system will detect that 192.0.2.110/24 and 192.0.2.100/24 are both on the 192.0.2.0/24 network and use the eth0 device because 192.0.2.100/24 is assigned to that device. If this scenario is not possible - you don't have the same network for the public addressees and node adresses - one can use the `destination` to specify a different network to match. In the imperative cluster create command that's something like `--public-address= 198.51.100.50/24%192.0.2.0/24` to tell it to use the device matching the 192.0.2.100/24 network (eth0 in our example above)? When I have some time I may try to use these examples in future doc updates. > BTW: there are IP ranges reserved for documentation: > https://datatracker.ietf.org/doc/html/rfc5737 > > "The blocks 192.0.2.0/24 (TEST-NET-1), 198.51.100.0/24 (TEST-NET-2), > and 203.0.113.0/24 (TEST-NET-3) are provided for use in > documentation." > Thanks for the reminder. I'm usually pretty good about properly using reserved dns names for documentation, but I forget when it comes to IPs and tend to just copy-paste from my actual test clusters. _______________________________________________ ceph-users mailing list -- [email protected] To unsubscribe send an email to [email protected]
