Snaplex Node Properties

Each Snap in a pipeline is estimated to consume one execution slot — the actual threads used can be higher, since Snaps like Pipeline Execute, bulk loaders, and Snaps performing input/output can use more than one thread each. A node only accepts a pipeline execution if its free slots are greater than or equal to the estimated slot count; otherwise the node rejects the execution and the scheduler tries another node in the Snaplex. If no node in the Snaplex has enough free slots, Scheduled Task and Ultra Task executions are placed in a Queued state until a node has capacity, while Triggered Task executions retry for one minute before returning a failure to the requesting client.

Default value: 4000

Refer to the Execution capacity section of Node properties for more detail.

The percentage of a node's Max. slots reserved for pipelines or validations executed interactively through Designer — typically during development and testing. If no slots are reserved, interactive executions can be blocked by long-running pipelines. Pipelines started via Tasks or the Pipeline Execute Snap do not have access to reserved slots.

Changes to this setting take effect immediately without a node restart.

Default value: 15

The percentage of a node's maximum JVM heap size that can be in use before the node is excluded from receiving new pipeline executions. Changes to this setting take effect immediately without a node restart.

Default value: 85

When a Snaplex node transitions into a cooldown state — for example, during a version update, a configuration change, a user-initiated restart, or when you place a node in maintenance mode — the node stops accepting new pipeline requests and waits up to this amount of time for its currently running pipelines to finish. If pipelines are still running when the limit is reached, they are automatically stopped so the node can complete its transition.

Enabling the Set restart wait time to Forever toggle disables the numeric field; the node then never automatically terminates a running pipeline during a transition — you're responsible for manually stopping pipelines so the transition can proceed. We recommend keeping this value under 60 minutes, since the control plane only waits up to an hour for nodes to update during mandatory upgrades before it stops sending them new pipeline requests.

Default value: 15 minutes

Refer to How Restart Maximum Wait Time works for complete details, including cooldown, maintenance mode, and version-change behavior.

The maximum JVM heap size for the node. The default, auto, has the Platform calculate the heap size automatically based on available system memory, leaving sufficient memory for operating system usage.

To set a custom value instead of auto, follow the number with a unit suffix with no space — for example, 5g for 5 GiB.

Default value: auto

Refer to Memory management for heap sizing guidance, compressed class space, and swap memory configuration. Groundplex only.

Controls which network interface the node accepts HTTP connections on:

• Localhost only (127.0.0.1) — restricts connections to the local machine
• Any interface (0.0.0.0) — accepts connections from any network interface

If neither preset fits — for example, the node needs to advertise a specific IP address or DNS hostname other nodes use to reach it — set the cc_secure_uri JVM option instead; refer to Set a specific IP address for a Groundplex node.

Default value: Localhost only (127.0.0.1)

Refer to the Network section of Node properties, and to Groundplex requirements: Network for firewall and port planning. Groundplex only.

The port on which the Snaplex node listens for HTTP connections.

Default value: 8090 for a JCC node and 8091 for a FeedMaster node.

Refer to HTTP port configuration for host port requirements, and JCC node port customization for how to change it. Groundplex only.

The port on which the Snaplex node listens for HTTPS connections. If left blank, the node falls back to its own internal default: port 8081 for a JCC node, 8084 for a FeedMaster node.

Default value: None.

Refer to JCC node port customization for how to change it. Groundplex only.

Maps hostnames to node roles (JCC or FeedMaster). By default, every node in a Groundplex runs as a JCC; use this field only if you want a specific machine to run as a FeedMaster instead. Add additional entries with + Add new node type.

This field is available only to Environments that have subscribed to Ultra Tasks, since FeedMaster is the node type that manages Ultra Task request distribution.

Default value: None.

Refer to the Node configuration section of Node properties. Groundplex only.

Key/value pairs for internal JVM and JCC configuration options. Do not edit these values without guidance from your CSM. Add additional entries with + Add new global property.

For example, if the FeedMaster doesn't receive an acknowledgment that a document was processed by a low-latency Ultra Task, it redelivers the document up to five times by default; you can customize this redelivery limit for a specific Snaplex by setting the ultra.max_redelivery_count property here. Headless Ultra Tasks don't support this property.

Default value: None.

Refer to Snaplex configuration for Snaps for more global property use cases. Groundplex only.