Multiple standalone Deep Discovery Analyzer appliances can be deployed and configured to form a cluster
that provides fault tolerance, improved performance, or a combination thereof.
Depending on your requirements and the number of Deep Discovery Analyzer appliances
available, you may deploy the following cluster configurations:
Cluster Configurations
Cluster Configuration
|
Description
|
||
High availability cluster
|
In a high availability cluster, one appliance acts as the
active primary appliance, and one acts as the passive primary appliance. The
passive primary appliance automatically takes over as the new active primary
appliance if the active primary appliance encounters an error and is unable to
recover.
|
||
Load-balancing cluster
|
In a load-balancing cluster, one appliance acts as the active
primary appliance, and any additional appliances act as secondary appliances. The
secondary appliances process submissions allocated by the active primary appliance
for performance improvement.
|
||
High availability cluster with load balancing
|
In a high availability cluster with load balancing, one
appliance acts as the active primary appliance, one acts as the passive primary
appliance, and any additional appliances act as secondary appliances. The passive
primary appliance takes over as the active primary appliance if the active primary
appliance encounters an error and is unable to recover. The secondary appliances
process submissions allocated by the active primary appliance for performance
improvement.
|
For details, see the Deep Discovery Analyzer Installation and
Deployment Guide.
The following table lists the available configuration modes and associated appliance
behavior.
Cluster Configuration Modes
Configuration Mode
|
Description
|
Primary (Active)
|
|
Primary (Passive)
|
|
Secondary
|
|
NoteIn environments that use a load-balancing cluster or a High Availability
cluster with load balancing, Deep Discovery Analyzer automatically slows down Virtual Analyzer throughput on the active
primary appliance to prevent exhaustion of system resources
|