Integrated source
cron3.0pl1-cl2.tar.gz
Vixie cron 3.0pl1 with the cluster changes already integrated, ready to build on Linux.
Download integrated source
What it does
Runs cron with an alternative root directory so crontabs, logs, and PID files can live inside a cluster package.
Why it mattered
Standard cron is system-wide. In failover clusters that meant jobs could run on nodes where the package was no longer present.
What to read first
Start with the README for the design rationale, the security checks around CRONDIR, and the build/install notes.
Downloads
Source patch and integrated tree
Patch
patch-cron3.0pl1-cl2.gz
The standalone patch against Vixie cron 3.0pl1 for anyone who wants the delta instead of the bundled source tree.
Download the patchDocumentation
README.CLUSTER
Design background, operational behavior, security checks, and the build/install outline for the cluster-aware daemon.
Read the READMENotes
Why standard cron was not enough
Operational problem
System-wide scheduling clashes with package failover
In HA environments, packages move between nodes. Standard cron does not know that, so jobs may continue running where the package has already left, creating failed runs, noisy logs, and messy monitoring.
Approach
Use a package-specific cron root
Cluster cron adds support for an alternative administration root. With -d or CRONDIR, cron stores its crontabs, logs, and PID file relative to the package filesystem.
Security
Guard rails around CRONDIR
The designated root must exist, be owned by root, not be writable by non-root users, and contain a cluster_cron.auth file owned by root.
History
The original project context
Cluster cron was framed as a better answer than wrapper scripts, monitoring exceptions, or fragile install/remove choreography during failover. The original README remains the authoritative explanation.