apache/beam · error · RuntimeError

not found. Please build the server with \n cd ; ./gradlew

Error message

%s not found. Please build the server with \n  cd %s; ./gradlew %s

What it means

path_to_beam_jar() normally locates a prebuilt Beam jar locally or falls back to a Maven repository. For a '.dev' (development/snapshot) version there is no remote artifact, so it immediately raises, telling you to build the jar yourself with gradlew and the given gradle target.

Solutions

1. Build the required jar: cd <beam-project-root>; ./gradlew <gradle_target> (printed in the error).
2. Alternatively pass the path to a prebuilt jar via the runner's jar option (e.g. --flink_job_server_jar or expansion service jar argument).
3. Or install a released Beam version (pip install apache-beam==X.Y.Z) whose jars exist on Maven Central.
4. For .dev use with remote execution, use the nightly Maven snapshots manually and point the server to that jar.

Example fix

// before
# beam dev version, no local jar
FlinkRunner()
// after
# cd ~/beam && ./gradlew :runners:flink:1.17:job-server:shadowJar
FlinkRunner(flink_job_server_jar='/home/me/beam/runners/flink/.../beam-runners-flink-1.17-job-server.jar')

import apache_beam as beam
from apache_beam.utils import subprocess_server
ver = beam.__version__
if '.dev' in ver:
    print('Dev version: build the jar locally with ./gradlew <target>')

When it happens

Trigger: Using a .dev0/.devN Beam Python version (e.g. installed from git or a nightly) and invoking code that needs a Java job-server jar (Flink, Spark, expansion service) without the jar present in the local beam build output.

Common situations: Developers running Beam from a source checkout or nightly wheels; missing 'python -m pip install -e sdks/python' + './gradlew :runners:flink:1.17:job-server:shadowJar' build step; version mismatch between Python dev version and available Maven artifacts.

Understand the failure class

Background: "File not found" and ENOENT errors: why libraries can't find a file that should exist — this error's family across 50 libraries.

Related errors

• Class name must not be empty
• Failed to build package from
• Java is not correctly installed in JAVA_HOME=
• Java must be installed on this system to use this…
• No proto encoding for PaneInfoCoder, always part of…

AI-assisted analysis of apache/beam@12126d8942 (2026-09-13). Data as JSON: /api/errors/0bfc43c1028ffb76. Report an issue: GitHub.

Appendix: source

Thrown at sdks/python/apache_beam/utils/subprocess_server.py:478

      return cls._BEAM_SERVICES.replacements[gradle_target]

    _, artifact_id = cls.parse_gradle_target(gradle_target, artifact_id)
    project_root = os.path.sep.join(
        os.path.abspath(__file__).split(os.path.sep)[:-5])
    local_path = cls.path_to_dev_beam_jar(
        gradle_target, appendix, version, artifact_id)
    if os.path.exists(local_path):
      _LOGGER.info('Using pre-built snapshot at %s', local_path)
      return local_path

    maven_repo = maven_repository_url or cls.MAVEN_CENTRAL_REPOSITORY
    if 'rc' in version:
      # Release candidate
      version = version.split('rc')[0]
      maven_repo = cls.MAVEN_STAGING_REPOSITORY
    elif '.dev' in version:
      # TODO: Attempt to use nightly snapshots?
      raise RuntimeError(
          (
              '%s not found. '
              'Please build the server with \n  cd %s; ./gradlew %s') %
          (local_path, os.path.abspath(project_root), gradle_target))

    return cls.path_to_maven_jar(
        artifact_id, cls.BEAM_GROUP_ID, version, maven_repo, appendix=appendix)

  @classmethod
  def _download_jar_to_cache(
      cls, download_url, cached_jar_path, user_agent=None):
    """Downloads a jar from the given URL to the specified cache path.
    
    Args:
      download_url (str): The URL to download from.
      cached_jar_path (str): The local path where the jar should be cached.
      user_agent (str): The user agent to use when downloading.
    """

View on GitHub (pinned to 12126d8942)