How can we limit the number of graphs running concurrently on CloverDX Server?

Resolved Cases
1

We have a requirement to restrict the number of graphs running in parallel on our CloverDX Server. Specifically, we want to limit concurrent executions to 30.

We’re aware of parameters like jobqueue.maxLoadStep, jobqueue.initLoadStep, and jobqueue.minLoadStep. Do we need to configure all of these to enforce the limit? Or is setting just jobqueue.maxLoadStep sufficient?

2

While the jobqueue.* parameters (initLoadStep, minLoadStep, maxLoadStep) are useful for controlling how quickly the server scales up job execution under load, they do not enforce a hard limit on the number of concurrently running graphs.

If your goal is to cap the total number of concurrent executions of the same job at 30, you’ll want to use a different set of parameters:

executor.max_running_concurrently = 30
executor.enqueue_executions = true

These settings should be added in the CloverDX Server Console under:
Configuration → Setup → Configuration File

  • executor.max_running_concurrently sets the global maximum number of concurrently running graphs.
  • executor.enqueue_executions = true ensures that any jobs exceeding the limit are queued instead of being rejected.

This setup will enforce the 30-job limit across all sandboxes and executions system-wide.

Optional: Sandbox-level limits

If you’d prefer to limit concurrency per sandbox or job, you can use the same properties (without the executor. prefix) for specific sanboxes or jobs. See our
documentation for more details.

Note:

  • These properties only limit the number of concurrent runs of the same job.
    They do not restrict the total number of all jobs running on CloverDX Server.
    It is not possible to globally limit the total number of concurrently running jobs across all sandboxes.
  • In a cluster environment, the limit is applied per node, not globally.