The HotStandby solution uses pairs of Primary and Secondary servers to provide true high availability. However, using pairs of servers might not be optimal for every use scenario. If near-instantaneous failover is not required, you might not be able to justify the expense of having a Secondary for every Primary server. At the other extreme, some business cases might need extra reliability and also warrant the resources to purchase spares for the spares. Spare for spares means that you might want to purchase not only a Secondary for every Primary, but also one or more spare servers so that when a Primary fails and its Secondary replaces it, a spare can be used as the new Secondary if the original Primary cannot be repaired quickly.
To help you to reduce costs or increase reliability, HotStandby supports various alternatives to the so called N+N or 2N standard hot standby model (in the standard model the number of Primary and Secondary servers is the same (N)). The alternatives include:
▪N + 1 Spare or N + M Spares: This is the Spare Node scenario for stand-alone servers. There are N “primary” servers and one or more spares. There are no Secondary servers. A failed (primary) server is replaced with a spare. The stand-alone server case is not a true hot standby scenario. It can be called warm standby: the spare computer is available but it does not have a copy of the database.
▪2N + 1 Spare or 2N + M Spares: This is the Spare Node scenario for HotStandby. There are N HotStandby pairs, that is, every Primary has a Secondary. In addition, there are M spares, where M is at least one and usually less than N. When a Primary or Secondary fails, a spare is brought in as the new Secondary. Thus a Primary server never operates alone for long, even if its original partner has failed.