General information about using a failover IP
The use of failover IP requires that you keep your data synchronous between both servers. There are basically different procedures for this, but this article deals only with the correct use of the failover IP.
All services that should be accessible via the failover IP must be configured to listen on the failover IP.
The failover IP is to be used as a so-called "Routed-IP". In Central Customer Management (ZKM), you can select a destination that corresponds to a server for each failover IP. The failover IP is routed so that the main IP of the selected server is the router responsible for this IP. This means that the failover IPs cannot be used for V servers operating in bridged mode (e.g. VMware). If the main IP of a server is not booted, the failover IP packets will not arrive.
In order to avoid an IP conflict when you run two servers in one segment, it is important that the IP is always booted on only one machine. In the event of an error, the IP can first be taken over by the new system and then booted. Ideally, the failed system is switched off via hardware KVM so that it does not restart and the IP restarts, which would then lead to an IP conflict.Tags: adresse, api, failover, general, ip