Pairing floating IPs for replication
Pairing floating IPs for replication
When you replicate NAS or object volumes to a remote StorONE system, clients need a stable address to reach whichever side is currently primary. Pairing a local floating IP with a remote one is how you set that up: the two floating IPs are linked, and only one side is ever active (holding the real address) at a time. If you haven't already created a floating IP to pair, see Floating IPs first.
The concepts specific to replication itself —— what continuous replication is, how failover works, and when a transition actually needs to happen —— are covered in the Replication chapter. This page only covers the floating IP side of that mechanism.
Pairing
floatingips pair links a local floating IP to one on a remote system. If a floating IP with the same name doesn't already exist on the remote system, StorONE creates one there automatically as an inactive placeholder (a "stub"), bound to the remote nodes and interfaces you specify —— it doesn't get a real address until it's activated with floatingips transition. A floating IP that's already in use for replication traffic itself can't be paired this way.
Using the web interface
- Open Inventory > Floating IPs, select the floating IP, and select Pair.
- Under Local Floating Ip, confirm the local floating IP.
- Under Target Floating Ip, choose the remote Target System to pair with.
- Under Target Interfaces per Node, add one row per remote node the floating IP might run on, choosing the network interface to use on that node.
- Confirm to create the pairing.
Using the CLI
floatingips pair <name> --targetSystem=<name> --remoteNodes=<text...> --remoteInterfaces=<text...>
| Option | Description |
|---|---|
<name> |
The local floating IP to pair. |
--targetSystem=<name> |
The remote system to pair with. |
--remoteNodes=<text...> |
The remote nodes to bind the floating IP to, paired positionally with --remoteInterfaces. |
--remoteInterfaces=<text...> |
The network interface on each corresponding remote node, paired positionally with --remoteNodes. |
Example
Pair the local nfs-vip floating IP with a stub on the dr-site system, bound to both of its nodes:
floatingips pair nfs-vip --targetSystem dr-site --remoteNodes dr-node-01,dr-node-02 --remoteInterfaces eth0,eth0
Unpairing
floatingips unpair removes the link between a local floating IP and its remote pair.
Using the web interface
- Open Inventory > Floating IPs, select the floating IP, and select Unpair.
- Under Local Floating Ip, confirm the local floating IP.
- Under Target Floating Ip System, choose the remote system to unpair from.
- Confirm.
Using the CLI
floatingips unpair <name> --targetSystem=<name> [--force]
| Option | Description |
|---|---|
<name> |
The local floating IP to unpair. |
--targetSystem=<name> |
The remote system to unpair from. |
--force |
Skip the confirmation prompt. |
Example
floatingips unpair nfs-vip --targetSystem dr-site
Activating a paired floating IP
floatingips transition activates a paired floating IP that's currently an inactive stub, giving it a real address and making it primary. This is the step that hands client connectivity over to this side of a replication pair —— normally as part of a failover to this system, described in the Replication chapter.
Using the web interface
- Open Inventory > Floating IPs and select Transition.
- Under Local Floating Ip, choose the inactive (stub) floating IP to activate.
- Confirm.
Using the CLI
floatingips transition --activateFloatingIp=<name>
| Option | Description |
|---|---|
--activateFloatingIp=<name> |
The inactive (stub) floating IP to activate. |
Example
floatingips transition --activateFloatingIp nfs-vip
See also: Floating IPs ·· Replication (failover and the broader replication pairing workflow)