High availability

When two or more BeyondCron servers are started, they automatically form a cluster, providing high availability as well as horizontal scaling.

Unlike some application clusters, BeyondCron clusters are masterless, remaining fully functional when a node fails, and automatically recovering when a failed node rejoins the cluster.

Standalone BeyondCron

When only one BeyondCron server is running, it operates in standalone mode.

1 single BeyondCron server.
2 BeyondCron agents, command line and web consoles.

If there is a hardware failure:

• job scheduling will cease;
• changes to job/calendar definitions cannot be made;
• agents will continue executing any already scheduled jobs, and wait for the server to restart;
• command line and web consoles will exit/logout.

Clustered BeyondCron servers

When multiple BeyondCron servers are running, they automatically form a cluster, providing:

• protection against hardware failure;
• horizontal scaling, ensuring that jobs remain on time as the BeyondCron environment grows.

1 multiple BeyondCron servers
2 BeyondCron agents, command line and web consoles.

Cluster node failure

When a cluster node crashes, or reboots:

• the remaining servers will automatically continue scheduling jobs;
• agents, command line and web consoles will remain connected to the cluster;
• the cluster remains fully operational allowing configuration changes to be made to jobs, calendars, etc…

Recovery – when the failed node restarts, it will automatically rejoin the cluster and recover any missed BeyondCron configuration changes.