NNMi Application Failover & failover time experiences
In the deployment guide HP talks for Application Failover about a failover time of 10-60 minutes to process the transactions logs of the embedded database. According to HP when using the Oracle database the failover time is almost instantaneous. The odd thing though is that when NNMi is running in a HA cluster (shared file system) they talk about a failover time of 5-30 minutes.
Why is this difference between failover time of a Oracle database based NNMi Application Failover setup and a NNMi HA cluster, anyone an idea?
Has anyone experiences with failover time in a NNMi Application Failover setup?
Re: NNMi Application Failover & failover time experiences
the Cluster solution (HA) based on a shared disk differs from the Application Level Failover (ALF) which must copy the data to the other server.
ALF times depend of the amount of data to transfer while HA needs time for mounting the shared disk and starting the processes.
My experience for ALF was around 15 minutes with 400 managed nodes (manual failover) - no HA experience until now.