getStreamGraphInfo

getStreamGraphInfo([name])

Return the metadata of the specified streaming graph.

If name is not specified, the function returns the metadata of all streaming graphs.

name (optional) is a string representing the name of the stream graph. You can provide either the fully qualified name (FQN), such as "catalog_name.orca_graph.graph_name", or just the graph name, like "factors". If only the name is given, the system will automatically complete it using the current catalog.

A table containing the following fields:

• id: Streaming graph ID.
• fqn: Fully qualified name of the streaming graph.
• graph: The structure of the streaming graph in JSON format, including:
  • version: Version number of the graph structure, used to prevent compatibility issues caused by version upgrades.
  • nodes: Nodes in the graph, each containing:
    • id: Node ID.
    • subgraphId: Subgraph ID the node belongs to.
    • taskid: Streaming task ID.
    • parallelism:
      • keyName: The column name used for partitioning parallel tasks.
      • count: Number of parallel tasks.
    • properties: Node properties.
    • inEdges: List of incoming edge IDs.
    • outEdges: List of outgoing edge IDs.
  • edges: Edges in the graph, each containing:
    • id: Edge ID.
    • inNodeId: ID of the upstream node.
    • outNodeId: ID of the downstream node.
    • partitionType:
      • FORWARD: Directly send upstream data to downstream.
      • NESTED: Distribute upstream data to downstream according to the nested relationship of parallelism.
      • SHUFFLE: Recombine upstream data into a single set and distribute to downstream based on parallelism.
    • filter: Filter condition on the edge.
    • handlers: List of functions to process upstream data; these functions will be merged into the handler parameter of subscribeTable.
  • config: Global configuration of the streaming graph.
• meta: Scheduling metadata of the streaming graph in JSON format, including:
  • id: Streaming graph ID.
  • status: Running status of the streaming graph:
    • building: Graph has been scheduled and is under construction.
    • running: Graph construction completed and running normally.
    • error: Recoverable fault encountered (e.g., node OOM); the system will reschedule the tasks.
    • failed: Unrecoverable fault encountered (e.g., user script error); the system will preserve the scene for later analysis.
    • destroying: Graph is being destroyed.
    • destroyed: Graph has been destroyed.
  • semantics: Consistency semantics, possible values:
    • exactly-once: Exactly-once execution.
    • at-least-once: At-least-once execution.
  • tasks: Scheduling information of streaming tasks.
  • createTime: Creation time.
  • reason: Reason for state transition.

getStreamGraphInfo("streamGraph1") // name is the name of the streaming graph
getStreamGraphInfo("catalog1.orca_graph.streamGraph1") // name is the fully qualified name