truefoundry 0.11.1rc1__tar.gz → 0.11.2__tar.gz

{truefoundry-0.11.1rc1 → truefoundry-0.11.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: truefoundry
-Version: 0.11.1rc1
+Version: 0.11.2
 Summary: TrueFoundry CLI
 Author-email: TrueFoundry Team <abhishek@truefoundry.com>
 Requires-Python: <3.14,>=3.8.1

{truefoundry-0.11.1rc1 → truefoundry-0.11.2}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "truefoundry"
-version = "0.11.1rc1"  # This is a placeholder version, it will be replaced when releasing
+version = "0.11.2"  # This is a placeholder version, it will be replaced when releasing
 requires-python = ">=3.8.1,<3.14"
 description = "TrueFoundry CLI"
 authors = [{ name = "TrueFoundry Team", email = "abhishek@truefoundry.com" }]

{truefoundry-0.11.1rc1 → truefoundry-0.11.2}/truefoundry/common/utils.py

@@ -16,6 +16,7 @@ from truefoundry.common.constants import (
     API_SERVER_RELATIVE_PATH,
     ENV_VARS,
     MLFOUNDRY_SERVER_RELATIVE_PATH,
+    TFY_API_KEY_ENV_KEY,
     TFY_DEBUG_ENV_KEY,
     TFY_HOST_ENV_KEY,
     TFY_INTERNAL_ENV_KEY,
@@ -113,9 +114,14 @@ def validate_tfy_host(tfy_host: str) -> None:
 def resolve_tfy_host(tfy_host: Optional[str] = None) -> str:
     tfy_host = tfy_host or ENV_VARS.TFY_HOST
     if not tfy_host:
-        raise ValueError(
-            f"Either `host` should be provided using `--host <value>`, or `{TFY_HOST_ENV_KEY}` env must be set"
-        )
+        if ENV_VARS.TFY_API_KEY:
+            raise ValueError(
+                f"TFY_HOST` env must be set since `{TFY_API_KEY_ENV_KEY}` env is set. Either set `{TFY_HOST_ENV_KEY}` or unset `{TFY_API_KEY_ENV_KEY}` and login"
+            )
+        else:
+            raise ValueError(
+                f"Either `host` should be provided using `--host <value>`, or `{TFY_HOST_ENV_KEY}` env must be set"
+            )
     tfy_host = tfy_host.strip("/")
     validate_tfy_host(tfy_host)
     return tfy_host

{truefoundry-0.11.1rc1 → truefoundry-0.11.2}/truefoundry/deploy/builder/builders/tfy_spark_buildpack/tfy_execute_notebook.py

@@ -45,16 +45,9 @@ def upload_file_to_s3(file_path, bucket_name, s3_key):
     # Use s3proxy for pushing data to s3
     # The JWT token is already available in the pod
     aws_access_key_id = os.environ.get("SPARK_APPLICATION_EVENT_LOG_JWT_TOKEN")
-    aws_secret_access_key = "__token__"
+    aws_secret_access_key = os.environ.get("TFY_NOTEBOOK_OUTPUT_S3_SECRET_KEY")
     s3_endpoint_url = os.environ.get("S3_PROXY_URL")
 
-    if not aws_access_key_id:
-        raise ValueError(
-            "SPARK_APPLICATION_EVENT_LOG_JWT_TOKEN environment variable is not set"
-        )
-    if not s3_endpoint_url:
-        raise ValueError("S3_PROXY_URL environment variable is not set")
-
     # Needed for the issue https://github.com/gaul/s3proxy/issues/765
     s3_config = Config(
         request_checksum_calculation="when_required",
@@ -106,7 +99,22 @@ def execute_notebook(notebook_path, output_path="/tmp/output.ipynb", parameters=
         stderr_file=sys.stderr,
     )
     print(f"Successfully executed notebook: {notebook_path}")
-    return output_path
+
+
+def validate_env_vars():
+    keys = [
+        "TFY_NOTEBOOK_OUTPUT_S3_KEY",
+        "TFY_NOTEBOOK_OUTPUT_S3_BUCKET",
+        "SPARK_APPLICATION_EVENT_LOG_JWT_TOKEN",
+        "TFY_NOTEBOOK_OUTPUT_S3_SECRET_KEY",
+    ]
+    unset_keys = [key for key in keys if not os.environ.get(key)]
+    if unset_keys:
+        raise ValueError(
+            f"Environment variables {unset_keys} are not set."
+            f"Contact you tenant-admin to configure storage bucket on the control plane "
+            f"to enable uploading spark notebook outputs."
+        )
 
 
 if __name__ == "__main__":
@@ -116,35 +124,31 @@ if __name__ == "__main__":
     parser.add_argument("notebook_path", help="Path to the notebook file to execute")
     args = parser.parse_args()
 
+    # Since failure to upload is considered a job failure, fail the job even before it run if uploads cannot happen
+    validate_env_vars()
+
     output_notebook_path = "/tmp/output.ipynb"
 
     # This would be the same as the default bucket used by servicefoundry-server
     s3_bucket = os.environ.get("TFY_NOTEBOOK_OUTPUT_S3_BUCKET")
-    # This would be something like sparkjob-events/<tenant-id>
-    s3_key_prefix = os.environ.get("TFY_NOTEBOOK_OUTPUT_S3_KEY_PREFIX")
+    # This would be something like sparkjob-events/<tenant-id>/output-notebooks/<application-id>/<jobrun-name>/output.html
+    s3_key = os.environ.get("TFY_NOTEBOOK_OUTPUT_S3_KEY")
 
     try:
-        executed_notebook_path = execute_notebook(
-            args.notebook_path, output_path=output_notebook_path
-        )
+        execute_notebook(args.notebook_path, output_path=output_notebook_path)
 
         # The following may also be modeled as an entrypoint
         # https://papermill.readthedocs.io/en/latest/extending-entry-points.html
         # Will take that up with next iteration where we save the executed notebook periodically
-        if s3_bucket and s3_key_prefix:
-            print("Converting notebook to HTML and uploading to S3...")
-            html_output_path = "/tmp/output.html"
-            convert_notebook_to_html(
-                notebook_path=executed_notebook_path, output_html_path=html_output_path
-            )
-
-            # Construct S3 key: use the original notebook name for the HTML file
-            notebook_name = os.path.basename(args.notebook_path)
-            s3_html_key = f"{s3_key_prefix}/output.html"
-            upload_file_to_s3(
-                file_path=html_output_path, bucket_name=s3_bucket, s3_key=s3_html_key
-            )
-            print(f"Successfully uploaded HTML to s3://{s3_bucket}/{s3_html_key}")
+        print("Converting notebook to HTML and uploading to S3...")
+        html_output_path = "/tmp/output.html"
+        convert_notebook_to_html(
+            notebook_path=output_notebook_path, output_html_path=html_output_path
+        )
+        upload_file_to_s3(
+            file_path=html_output_path, bucket_name=s3_bucket, s3_key=s3_key
+        )
+        print(f"Successfully uploaded HTML to s3://{s3_bucket}/{s3_key}")
 
     except Exception as e:
         print(f"Error executing notebook {args.notebook_path}: {e}")

{truefoundry-0.11.1rc1 → truefoundry-0.11.2}/truefoundry/ml/artifact/truefoundry_artifact_repo.py

@@ -6,17 +6,7 @@ import uuid
 from concurrent.futures import FIRST_EXCEPTION, Future, ThreadPoolExecutor, wait
 from shutil import rmtree
 from threading import Event
-from typing import (
-    Any,
-    Callable,
-    Dict,
-    Iterator,
-    List,
-    Optional,
-    Sequence,
-    Tuple,
-    Union,
-)
+from typing import Any, Callable, Dict, Iterator, List, Optional, Sequence, Tuple
 from urllib.parse import unquote
 from urllib.request import pathname2url
 
@@ -31,6 +21,7 @@ from rich.progress import (
 )
 from tqdm.utils import CallbackIOWrapper
 from truefoundry_sdk import (
+    FileInfo,
     MultiPartUploadResponse,
     MultiPartUploadStorageProvider,
     Operation,
@@ -52,11 +43,6 @@ from truefoundry.common.storage_provider_utils import (
 )
 from truefoundry.ml._autogen.client import (  # type: ignore[attr-defined]
     ApiClient,
-    FileInfoDto,
-    ListFilesForArtifactVersionRequestDto,
-    ListFilesForArtifactVersionsResponseDto,
-    ListFilesForDatasetRequestDto,
-    ListFilesForDatasetResponseDto,
     MlfoundryArtifactsApi,
     RunArtifactsApi,
 )
@@ -592,44 +578,23 @@ class MlFoundryArtifactsRepository:
             progress=progress,
         )
 
-    def _list_files(
-        self, artifact_identifier: ArtifactIdentifier, path, page_size, page_token
-    ) -> Union[ListFilesForDatasetResponseDto, ListFilesForArtifactVersionsResponseDto]:
-        if artifact_identifier.dataset_fqn:
-            return self._mlfoundry_artifacts_api.list_files_for_dataset_post(
-                list_files_for_dataset_request_dto=ListFilesForDatasetRequestDto(
-                    dataset_fqn=artifact_identifier.dataset_fqn,
-                    path=path,
-                    max_results=page_size,
-                    page_token=page_token,
-                )
-            )
-        else:
-            return self._mlfoundry_artifacts_api.list_files_for_artifact_version_post(
-                list_files_for_artifact_version_request_dto=ListFilesForArtifactVersionRequestDto(
-                    id=str(artifact_identifier.artifact_version_id),
-                    path=path,
-                    max_results=page_size,
-                    page_token=page_token,
-                )
-            )
-
     def list_artifacts(
         self, path=None, page_size=_LIST_FILES_PAGE_SIZE, **kwargs
-    ) -> Iterator[FileInfoDto]:
-        page_token = None
-        started = False
-        while not started or page_token is not None:
-            started = True
-            page = self._list_files(
-                artifact_identifier=self.artifact_identifier,
+    ) -> Iterator[FileInfo]:
+        if self.artifact_identifier.dataset_id:
+            for file_info in client.data_directories.list_files(
+                id=str(self.artifact_identifier.dataset_id),
                 path=path,
-                page_size=page_size,
-                page_token=page_token,
-            )
-            for file_info in page.files:
+                limit=page_size,
+            ):
+                yield file_info
+        else:
+            for file_info in client.artifact_versions.list_files(
+                id=str(self.artifact_identifier.artifact_version_id),
+                path=path,
+                limit=page_size,
+            ):
                 yield file_info
-            page_token = page.next_page_token
 
     def _is_directory(self, artifact_path):
         # TODO: Ideally server should return a flag to indicate if it is a directory