Explicit graceful shutdown for Spofire Server

You can manually stop Spotfire Server instances in a manner that controls the minimum draining and the draining timeout when the server needs to be restarted due to configuration changes.

When Spotfire Server has a configuration change, a software upgrade, or a change made to a schema that requires restarting, then you can explicitly shut down the server gracefully, controlling the following aspects of the shutdown, and informing the back-end services that the server instance is offline and should not be used for new requests.

Note: You can set up a URL to monitor the health status of load-balanced servers in a cluster. For more information about enabling this feature, see Enabling health check URL for load balanced servers.

The command graceful-shutdown accepts two values.


Setting	Description
`minimum-draining-seconds`	Specifies the minimum amount of time that the command waits before it returns.
`draining-timeout-seconds`	Specifies the maximum time that the command waits for the instance becoming idle. It must accommodate the longest operation (for example, an information services query or a save to library operation).

Example

graceful-shutdown --minimum-draining-seconds=2000 --draining-timeout-seconds=3600

Note: This functionality controls draining and shutting down only. The system relies on an external action to start the servers again; for example, using the systemd/Windows service or with Kubernetes.

If you do not explicitly call graceful-shutdown, and the server detects a configuration change that requires a restart, then the server stops. For information about configuring the server or the server cluster to always manage automatic shutdowns gracefully, see Graceful shutdown configuration settings.