17417: Merge branch 'main' into 17417-add-arm64

[arvados.git] / sdk / cwl / arvados_cwl / arvdocker.py

diff --git a/sdk/cwl/arvados_cwl/arvdocker.py b/sdk/cwl/arvados_cwl/arvdocker.py
index 0513ca02ec68ca2d0fe5f58ced9ab1c98c124844..26408317cbe6d0cdb5382f38c1455b0bd7b5db2a 100644 (file)
--- a/sdk/cwl/arvados_cwl/arvdocker.py
+++ b/sdk/cwl/arvados_cwl/arvdocker.py
@@ -18,9 +18,13 @@ logger = logging.getLogger('arvados.cwl-runner')
 cached_lookups = {}
 cached_lookups_lock = threading.Lock()
 
-def arv_docker_get_image(api_client, dockerRequirement, pull_image, project_uuid):
+def arv_docker_get_image(api_client, dockerRequirement, pull_image, project_uuid,
+                         force_pull, tmp_outdir_prefix):
     """Check if a Docker image is available in Keep, if not, upload it using arv-keepdocker."""
 
+    if "http://arvados.org/cwl#dockerCollectionPDH" in dockerRequirement:
+        return dockerRequirement["http://arvados.org/cwl#dockerCollectionPDH"]
+
     if "dockerImageId" not in dockerRequirement and "dockerPull" in dockerRequirement:
         dockerRequirement = copy.deepcopy(dockerRequirement)
         dockerRequirement["dockerImageId"] = dockerRequirement["dockerPull"]
@@ -31,9 +35,9 @@ def arv_docker_get_image(api_client, dockerRequirement, pull_image, project_uuid
     global cached_lookups_lock
     with cached_lookups_lock:
         if dockerRequirement["dockerImageId"] in cached_lookups:
-            return dockerRequirement["dockerImageId"]
+            return cached_lookups[dockerRequirement["dockerImageId"]]
 
-    with SourceLine(dockerRequirement, "dockerImageId", WorkflowException):
+    with SourceLine(dockerRequirement, "dockerImageId", WorkflowException, logger.isEnabledFor(logging.DEBUG)):
         sp = dockerRequirement["dockerImageId"].split(":")
         image_name = sp[0]
         image_tag = sp[1] if len(sp) > 1 else "latest"
@@ -45,7 +49,10 @@ def arv_docker_get_image(api_client, dockerRequirement, pull_image, project_uuid
         if not images:
             # Fetch Docker image if necessary.
             try:
-                cwltool.docker.get_image(dockerRequirement, pull_image)
+                result = cwltool.docker.DockerCommandLineJob.get_image(dockerRequirement, pull_image,
+                                                              force_pull, tmp_outdir_prefix)
+                if not result:
+                    raise WorkflowException("Docker image '%s' not available" % dockerRequirement["dockerImageId"])
             except OSError as e:
                 raise WorkflowException("While trying to get Docker image '%s', failed to execute 'docker': %s" % (dockerRequirement["dockerImageId"], e))
 
@@ -57,8 +64,10 @@ def arv_docker_get_image(api_client, dockerRequirement, pull_image, project_uuid
             args.append(image_tag)
             logger.info("Uploading Docker image %s:%s", image_name, image_tag)
             try:
-                arvados.commands.keepdocker.main(args, stdout=sys.stderr)
+                arvados.commands.put.api_client = api_client
+                arvados.commands.keepdocker.main(args, stdout=sys.stderr, install_sig_handlers=False, api=api_client)
             except SystemExit as e:
+                # If e.code is None or zero, then keepdocker exited normally and we can continue
                 if e.code:
                     raise WorkflowException("keepdocker exited with code %s" % e.code)
 
@@ -69,10 +78,12 @@ def arv_docker_get_image(api_client, dockerRequirement, pull_image, project_uuid
         if not images:
             raise WorkflowException("Could not find Docker image %s:%s" % (image_name, image_tag))
 
+        pdh = api_client.collections().get(uuid=images[0][0]).execute()["portable_data_hash"]
+
         with cached_lookups_lock:
-            cached_lookups[dockerRequirement["dockerImageId"]] = True
+            cached_lookups[dockerRequirement["dockerImageId"]] = pdh
 
-    return dockerRequirement["dockerImageId"]
+    return pdh
 
 def arv_docker_clear_cache():
     global cached_lookups