When two or more BeyondCron servers are started, they automatically form a cluster, providing not only high availability, but also horizontal scaling. Horizontal scaling ensures that BeyondCron does not no become CPU or memory constrained as the job list grows.

Standalone BeyondCron

When only one BeyondCron server is running, all job definitions, scheduling and agent coordination are managed by a single server. Apart from being a single point of failure, this has the potential to limit the size of your BeyondCron environment due to CPU and/or memory limits.

Standalone server

1 job memory stack.

2-way cluster

When a 2-way cluster is formed, BeyondCron will:

  • automatically distribute the job definitions between the 2 servers (A & B);
  • create backup copies of the job definitions on the alternate servers (A’ & B’) for high availability recovery;
  • schedule jobs from both servers.

2 way cluster

A BeyondCron server A.
B BeyondCron server B.

whilst a 2-way cluster provides increased CPU capacity, any memory limits remain unchanged due to memory backups used for high availability recovery.

N-way cluster

When 3-way or greater cluster is formed, memory utililisation on each node will also decrease, as job definitions are distributed across more nodes.

3 way cluster

A BeyondCron server A.
B BeyondCron server B.
C BeyondCron server C.