All Products
Search

This Product

    CloudFlow:Iteration (Map)

    Document Center

    CloudFlow:Iteration (Map)

    Last Updated:Jun 23, 2026

    A Map state executes the same operation on each element in an array, simplifying large-scale data processing and workflow management.

    Processing modes

    The Map state supports two processing modes:

    • Inline mode

      Each iteration runs in the context of the parent workflow. By default, a Map state uses inline mode. For more information, see Inline mode.

    • Distributed mode

      Each iteration runs as a child workflow, supporting up to 10,000 iterations. For more information, see Distributed mode.

    Differences between inline and distributed modes

    Comparison criteria

    Inline mode

    Distributed mode

    Data source

    Reads data from the previous state in the workflow.

    • Reads data from the previous state in the workflow.

    • Reads a list of objects from an Object Storage Service (OSS) bucket.

    • Reads a CSV file from an OSS bucket.

    • Reads an inventory from an OSS bucket.

    • Reads a JSON file from an OSS bucket.

    Iteration process

    Each iteration runs in the context of the parent workflow.

    Each iteration runs as a child workflow, supporting up to 10,000 iterations.

    Maximum concurrency

    The default maximum concurrency is 40.

    The default maximum concurrency is 100. To request a quota increase, submit a ticket.

    Maximum number of iterations

    No limit on the number of iterations by default. However, the total iteration output size is limited by the Event IO Size.

    The default maximum is 10,000 iterations. To increase the quota up to 1,000,000, submit a ticket.

    Payload

    Input payload size is limited to 64 KB.

    The 64 KB input payload limit does not apply when you read external input through an ItemReader.