Migrate gsutil usage to gcloud storage (#4292)

* Migrate gsutil usage to gcloud storage

* Manual change

* update

* Update NotebookProcessors.py

* Update NotebookProcessors.py

---------

Co-authored-by: gurusai-voleti <gvoleti@google.com>

.cloud-build/execute_notebook_helper.py

@@ -56,8 +56,8 @@ def execute_notebook(
     print("\n=== DOWNLOAD EXECUTED NOTEBOOK ===\n")
     print(f"Please debug the executed notebook by downloading the executed notebook:")
 
-    print("Option 1. Using gsutil. Run the following command in your terminal.")
-    print(f'\tgsutil cp "{output_file_or_uri}" .')
+    print("Option 1. Using gcloud storage. Run the following command in your terminal.")
+    print(f'\tgcloud storage cp "{output_file_or_uri}" .')
 
     print("Option 2. Using this link.")
     print(f"\thttps://storage.googleapis.com/{output_file_or_uri[5:]}")

.cloud-build/utils/NotebookProcessors.py

@@ -108,7 +108,7 @@ class VertexAIInstallProprocessor(Preprocessor):
         if "google-cloud-aiplatform" not in content:
             return content
         return (
-            f"gsutil cp {self.vertex_ai_wheel} google-cloud-aiplatform.whl\n" + 
+            f"gcloud storage cp {self.vertex_ai_wheel} google-cloud-aiplatform.whl\n" + 
             content.replace("google-cloud-aiplatform\n", "google-cloud-aiplatform.whl\n")
             .replace("google-cloud-aiplatform ", "google-cloud-aiplatform.whl ")
         )

.cloud-build/utils/util.py

@@ -15,7 +15,7 @@ def download_file(bucket_name: str, blob_name: str, destination_file: str) -> st
     remote_file_path = "".join(["gs://", "/".join([bucket_name, blob_name])])
 
     subprocess.check_output(
-        ["gsutil", "cp", remote_file_path, destination_file], encoding="UTF-8"
+        ["gcloud", "storage", "cp", remote_file_path, destination_file], encoding="UTF-8"
     )
 
     return destination_file
@@ -27,7 +27,7 @@ def upload_file(
 ) -> str:
     """Copies a local file to a GCS path"""
     subprocess.check_output(
-        ["gsutil", "cp", local_file_path, remote_file_path], encoding="UTF-8"
+        ["gcloud", "storage", "cp", local_file_path, remote_file_path], encoding="UTF-8"
     )
 
     return remote_file_path

community-content/pipeline_components/google-cloud/storage/download/component.yaml

@@ -24,12 +24,12 @@ implementation:
 
             # Checking whether the URI points to a single blob, a directory or a URI pattern
             # URI points to a blob when that URI does not end with slash and listing that URI only yields the same URI
-            if [[ "$uri" != */ ]] && (gsutil ls "$uri" | grep --fixed-strings --line-regexp "$uri"); then
+            if [[ "$uri" != */ ]] && (gcloud storage ls "$uri" | grep --fixed-strings --line-regexp "$uri"); then
                 mkdir -p "$(dirname "$output_path")"
-                gsutil -m cp -r "$uri" "$output_path"
+                gcloud storage cp --recursive "$uri" "$output_path"
             else
                 mkdir -p "$output_path" # When source path is a directory, gsutil requires the destination to also be a directory
-                gsutil -m rsync -r "$uri" "$output_path" # gsutil cp has different path handling than Linux cp. It always puts the source directory (name) inside the destination directory. gsutil rsync does not have that problem.
+                gcloud storage rsync --recursive "$uri" "$output_path" # gsutil cp has different path handling than Linux cp. It always puts the source directory (name) inside the destination directory. gsutil rsync does not have that problem.
             fi
         - inputValue: GCS path
         - outputPath: Data

community-content/pytorch_text_classification_using_vertex_sdk_and_gcloud/custom_container/scripts/train-cloud.sh

@@ -77,4 +77,4 @@ echo "After the job is completed successfully, model files will be saved at $JOB
 
 # # Verify the model was exported
 # echo "Verify the model was exported:"
-# gsutil ls ${JOB_DIR}/
+# gcloud storage ls ${JOB_DIR}/

community-content/pytorch_text_classification_using_vertex_sdk_and_gcloud/predictor/Dockerfile.serve

@@ -34,4 +34,4 @@ RUN echo "service_envelope=json\n"     "inference_address=http://0.0.0.0:${AIP_H
 USER model-server
 
 # run Torchserve HTTP serve to respond to prediction requests
-CMD ["echo", "AIP_STORAGE_URI=${AIP_STORAGE_URI}", ";",     "gsutil", "cp", "-r", "${AIP_STORAGE_URI}/${MODEL_NAME}.mar", "/home/model-server/model-store/", ";",     "ls", "-ltr", "/home/model-server/model-store/", ";",     "torchserve", "--start", "--ts-config=/home/model-server/config.properties",     "--models", "${MODEL_NAME}=${MODEL_NAME}.mar",     "--model-store", "/home/model-server/model-store"]
+CMD ["echo", "AIP_STORAGE_URI=${AIP_STORAGE_URI}", ";",     "gcloud", "storage", "cp", "--recursive", "${AIP_STORAGE_URI}/${MODEL_NAME}.mar", "/home/model-server/model-store/", ";",     "ls", "-ltr", "/home/model-server/model-store/", ";",     "torchserve", "--start", "--ts-config=/home/model-server/config.properties",     "--models", "${MODEL_NAME}=${MODEL_NAME}.mar",     "--model-store", "/home/model-server/model-store"]

community-content/pytorch_text_classification_using_vertex_sdk_and_gcloud/python_package/scripts/train-cloud.sh

@@ -67,4 +67,4 @@ echo "After the job is completed successfully, model files will be saved at $JOB
 
 # # Verify the model was exported
 # echo "Verify the model was exported:"
-# gsutil ls ${JOB_DIR}/
+# gcloud storage ls ${JOB_DIR}/

community-content/vertex_model_garden/model_oss/cloudnerf/render.sh

@@ -99,14 +99,14 @@ create_dir_if_not_exists "$CHECKPOINTS_PATH"
 touch "$local_experiment_path/$exp_folder_name/log_render.txt"
 
 # Copy experiment from GCS bucket to local
-gsutil -m cp -r "${args[-gcs_experiment_path]}/data" "$local_experiment_path/$exp_folder_name" || exit 1
-gsutil -m cp -r "${args[-gcs_experiment_path]}/checkpoints/${training_job_name}/*" "$CHECKPOINTS_PATH" || exit 1
+gcloud storage cp --recursive "${args[-gcs_experiment_path]}/data" "$local_experiment_path/$exp_folder_name" || exit 1
+gcloud storage cp --recursive "${args[-gcs_experiment_path]}/checkpoints/${training_job_name}/*" "$CHECKPOINTS_PATH" || exit 1
 
 # Check and copy keyframes file.
 if [[ -n ${args[-gcs_keyframes_file]} ]]; then
   keyframes_file_basename=$(basename "${args[-gcs_keyframes_file]}")
   local_keyframes_file="$local_dataset_path/$keyframes_file_basename"
-  gsutil cp "${args[-gcs_keyframes_file]}" "$local_keyframes_file" || exit 1
+  gcloud storage cp "${args[-gcs_keyframes_file]}" "$local_keyframes_file" || exit 1
   echo "Local keyframe file: $local_keyframes_file"
   launch_rendering "$local_keyframes_file"
 else
@@ -114,4 +114,4 @@ else
 fi
 
 # Copy rendered data back to GCS.
-gsutil -m cp -r "$OUTPUT_RENDER_PATH" "${args[-gcs_experiment_path]}/render/${rendering_job_name}"
+gcloud storage cp --recursive "$OUTPUT_RENDER_PATH" "${args[-gcs_experiment_path]}/render/${rendering_job_name}"

community-content/vertex_model_garden/model_oss/cloudnerf/train.sh

@@ -74,7 +74,7 @@ create_dir_if_not_exists "$local_experiment_path"
 create_dir_if_not_exists "$local_experiment_path/$scene_folder_name"
 
 # Copy experiment from GCS bucket to local.
-gsutil -m cp -r "${gcs_experiment_path}/data" "$local_experiment_path/$scene_folder_name" || exit 1
+gcloud storage cp --recursive "${gcs_experiment_path}/data" "$local_experiment_path/$scene_folder_name" || exit 1
 
 echo "GCS Experiment: $gcs_experiment_path"
 echo "Gin Config File: $gin_config_file"
@@ -89,6 +89,6 @@ accelerate launch train.py --gin_configs="$gin_config_file" \
   --gin_bindings="Config.factor = ${factor}" \
   --gin_bindings="Config.max_steps = ${max_training_steps}"
 
-gsutil -m rm -r "${gcs_experiment_path}/checkpoints/${training_job_name}"
-gsutil -m cp -r "$local_experiment_path/$scene_folder_name/config.gin" "${gcs_experiment_path}/${training_job_name}_config.gin"
-gsutil -m cp -r "$local_experiment_path/$scene_folder_name/checkpoints/*/*" "${gcs_experiment_path}/checkpoints/${training_job_name}"
+gcloud storage rm --recursive "${gcs_experiment_path}/checkpoints/${training_job_name}"
+gcloud storage cp --recursive "$local_experiment_path/$scene_folder_name/config.gin" "${gcs_experiment_path}/${training_job_name}_config.gin"
+gcloud storage cp --recursive "$local_experiment_path/$scene_folder_name/checkpoints/*/*" "${gcs_experiment_path}/checkpoints/${training_job_name}"

notebooks/community/ml_ops/setup.py

@@ -89,7 +89,7 @@ print("UUID", UUID)
 if args.bucket_required:
     BUCKET_NAME = PROJECT_ID + "aip-" + UUID
     BUCKET_URI = f"gs://{BUCKET_NAME}"
-    os.system(f"gsutil mb -l {REGION} {BUCKET_URI}")
+    os.system(f"gcloud storage buckets create --location={REGION} {BUCKET_URI}")
     print("BUCKET_URI", BUCKET_URI)