What is fault resilience?
SteelEye uses the term fault-resilient to describe a computing environment where servers can failover to other servers. Specifically, the term identifies environments where other servers take over for failed servers or failed applications providing up to 99.99% availability for systems and applications. Fault Resilience differs from Fault Tolerance in that Fault Resilience provides for a certain amount of accepted downtime during the failover process - generally an average of no more than 53 minutes a year. LifeKeeper provides fault resilience for Linux environments by enabling other servers in a cluster to take over for failed servers or failed applications.
Does LifeKeeper require dedicated hot back-up servers?
No. LifeKeeper supports an active-active server configuration. This configuration eliminates the need for extra servers dedicated for hot backup and allows clients and applications to failover to other production servers in the cluster.
What kinds of applications can LifeKeeper support?
LifeKeeper is an ideal fit for mission-critical applications. LifeKeeper can support most applications running in the Linux environment, including e-mail, databases, printers, application servers and Web servers.
Do applications need to be modified for use with LifeKeeper?
No. Depending on the type of application, the location of certain data that needs to be protected, will be denoted in LifeKeeper configuration guidelines. For example, if LifeKeeper protects the Sendmail application, the e-mail accounts database and mail spool directories would be stored on a shared device.
Can LifeKeeper protect applications from a single point of failure?
Yes. LifeKeeper eliminates a single point of failure for applications by enabling applications to failover to other servers in the cluster. This functionality not only minimizes the risk of a single point of failure, but also allows Linux systems to meet the stringent availability requirements of mission-critical operations by creating a fault resilient environment.
What is the impact on users when an application fails over to another server in the cluster?
The switchover that occurs, if an event creates an interruption in a server or application's availability, is transparent to end-users. LifeKeeper migrates all applications and transfers connectivity in such a way that clients have continuous access to applications and data. However, in configurations where a very large database application fails over, users may notice a few seconds of non-connectivity.
How many systems are required in a basic LifeKeeper configuration?
The minimum number of systems in a LifeKeeper configuration is two.
How much memory is required for each system in a LifeKeeper configuration?
The minimum amount of memory required for a LifeKeeper configuration is 32 MB, but 48 MB is recommended.