🐛Maintenance: Unlock botocore again 🚨🚨🚨🚨 (#7589)

Author: sanderegg

packages/aws-library/requirements/_base.txt

@@ -384,7 +384,7 @@ typer==0.15.2
     # via
     #   -r requirements/../../../packages/service-library/requirements/../../../packages/settings-library/requirements/_base.in
     #   -r requirements/../../../packages/settings-library/requirements/_base.in
-types-aiobotocore==2.21.0
+types-aiobotocore==2.21.1
     # via -r requirements/_base.in
 types-aiobotocore-ec2==2.21.0
     # via types-aiobotocore

packages/aws-library/requirements/_test.txt

@@ -280,17 +280,17 @@ sympy==1.13.3
     # via cfn-lint
 termcolor==2.5.0
     # via pytest-sugar
-types-aioboto3==14.0.0
+types-aioboto3==14.1.0
     # via -r requirements/_test.in
-types-aiobotocore==2.21.0
+types-aiobotocore==2.21.1
     # via
     #   -c requirements/_base.txt
     #   types-aioboto3
 types-awscrt==0.23.10
     # via
     #   -c requirements/_base.txt
     #   botocore-stubs
-types-boto3==1.37.4
+types-boto3==1.38.2
     # via -r requirements/_test.in
 types-s3transfer==0.11.3
     # via

packages/aws-library/src/aws_library/s3/__init__.py

@@ -22,10 +22,10 @@
 )
 
 __all__: tuple[str, ...] = (
-    "PRESIGNED_LINK_MAX_SIZE",
-    "S3_MAX_FILE_SIZE",
     "CopiedBytesTransferredCallback",
     "MultiPartUploadLinks",
+    "PRESIGNED_LINK_MAX_SIZE",
+    "S3_MAX_FILE_SIZE",
     "S3AccessError",
     "S3BucketInvalidError",
     "S3DestinationNotEmptyError",
@@ -37,8 +37,8 @@
     "S3RuntimeError",
     "S3UploadNotFoundError",
     "SimcoreS3API",
-    "UploadID",
     "UploadedBytesTransferredCallback",
+    "UploadID",
 )
 
 # nopycln: file

packages/aws-library/src/aws_library/s3/_client.py

@@ -3,7 +3,6 @@
 import functools
 import logging
 import urllib.parse
-import warnings
 from collections.abc import AsyncGenerator, Sequence
 from dataclasses import dataclass, field
 from pathlib import Path
@@ -12,7 +11,6 @@
 import aioboto3
 from aiobotocore.session import ClientCreatorContext
 from boto3.s3.transfer import TransferConfig
-from botocore import __version__ as botocore_version
 from botocore import exceptions as botocore_exc
 from botocore.client import Config
 from models_library.api_schemas_storage.storage_schemas import (
@@ -22,7 +20,6 @@
 )
 from models_library.basic_types import SHA256Str
 from models_library.bytes_iters import BytesIter, DataSize
-from packaging import version
 from pydantic import AnyUrl, ByteSize, TypeAdapter
 from servicelib.bytes_iters import DEFAULT_READ_CHUNK_SIZE, BytesStreamer
 from servicelib.logging_utils import log_catch, log_context
@@ -54,22 +51,6 @@
 )
 from ._utils import compute_num_file_chunks, create_final_prefix
 
-_BOTOCORE_VERSION: Final[version.Version] = version.parse(botocore_version)
-_MAX_BOTOCORE_VERSION_COMPATIBLE_WITH_CEPH_S3: Final[version.Version] = version.parse(
-    "1.36.0"
-)
-
-
-def _check_botocore_version() -> None:
-    if _BOTOCORE_VERSION >= _MAX_BOTOCORE_VERSION_COMPATIBLE_WITH_CEPH_S3:
-        warnings.warn(
-            f"Botocore version {botocore_version} is not supported for file uploads with CEPH S3 until CEPH is updated. "
-            "Please use a version < 1.36.0. The upload operation will likely fail.",
-            RuntimeWarning,
-            stacklevel=2,
-        )
-
-
 _logger = logging.getLogger(__name__)
 
 _S3_MAX_CONCURRENCY_DEFAULT: Final[int] = 10
@@ -107,13 +88,21 @@ async def create(
         session_client = None
         exit_stack = contextlib.AsyncExitStack()
         try:
+            config = Config(
+                # This setting tells the S3 client to only calculate checksums when explicitly required
+                # by the operation. This avoids unnecessary checksum calculations for operations that
+                # don't need them, improving performance.
+                # See: https://boto3.amazonaws.com/v1/documentation/api/latest/guide/s3.html#calculating-checksums
+                signature_version="s3v4",
+                request_checksum_calculation="when_required",  # type: ignore[call-arg]
+            )
             session_client = session.client(  # type: ignore[call-overload]
                 "s3",
                 endpoint_url=f"{settings.S3_ENDPOINT}",
                 aws_access_key_id=settings.S3_ACCESS_KEY,
                 aws_secret_access_key=settings.S3_SECRET_KEY,
                 region_name=settings.S3_REGION,
-                config=Config(signature_version="s3v4"),
+                config=config,
             )
             assert isinstance(session_client, ClientCreatorContext)  # nosec
 
@@ -523,9 +512,6 @@ async def upload_file(
         bytes_transfered_cb: UploadedBytesTransferredCallback | None,
     ) -> None:
         """upload a file using aioboto3 transfer manager (e.g. works >5Gb and creates multiple threads)"""
-
-        _check_botocore_version()
-
         upload_options: dict[str, Any] = {
             "Bucket": bucket,
             "Key": object_key,
@@ -550,9 +536,6 @@ async def copy_object(
         object_metadata: S3MetaData | None = None,
     ) -> None:
         """copy a file in S3 using aioboto3 transfer manager (e.g. works >5Gb and creates multiple threads)"""
-
-        _check_botocore_version()
-
         copy_options: dict[str, Any] = {
             "CopySource": {"Bucket": bucket, "Key": src_object_key},
             "Bucket": bucket,
@@ -659,7 +642,6 @@ async def upload_object_from_file_like(
         file_like_reader: FileLikeReader,
     ) -> None:
         """streams write an object in S3 from an AsyncIterable[bytes]"""
-        _check_botocore_version()
         await self._client.upload_fileobj(file_like_reader, bucket_name, object_key)  # type: ignore[arg-type]
 
     @staticmethod

services/dask-sidecar/requirements/_base.txt

@@ -1,6 +1,6 @@
 aio-pika==9.5.3
     # via -r requirements/../../../packages/service-library/requirements/_base.in
-aiobotocore==2.17.0
+aiobotocore==2.21.1
     # via s3fs
 aiocache==0.12.3
     # via -r requirements/../../../packages/service-library/requirements/_base.in
@@ -67,10 +67,8 @@ blosc==1.11.2
     # via -r requirements/_base.in
 bokeh==3.6.2
     # via dask
-botocore==1.35.93
-    # via
-    #   -c requirements/constraints.txt
-    #   aiobotocore
+botocore==1.37.1
+    # via aiobotocore
 certifi==2024.8.30
     # via
     #   -c requirements/../../../packages/common-library/requirements/../../../requirements/constraints.txt
@@ -569,7 +567,6 @@ urllib3==2.2.3
     #   -c requirements/../../../packages/settings-library/requirements/../../../packages/common-library/requirements/../../../requirements/constraints.txt
     #   -c requirements/../../../packages/settings-library/requirements/../../../requirements/constraints.txt
     #   -c requirements/../../../requirements/constraints.txt
-    #   aiobotocore
     #   botocore
     #   distributed
     #   requests

services/dask-sidecar/requirements/_test.txt

@@ -15,11 +15,11 @@ aws-xray-sdk==2.14.0
     # via moto
 blinker==1.9.0
     # via flask
-boto3==1.35.93
+boto3==1.37.1
     # via
     #   aws-sam-translator
     #   moto
-botocore==1.35.93
+botocore==1.37.1
     # via
     #   -c requirements/_base.txt
     #   aws-xray-sdk
@@ -226,7 +226,7 @@ rpds-py==0.22.3
     #   -c requirements/_base.txt
     #   jsonschema
     #   referencing
-s3transfer==0.10.4
+s3transfer==0.11.3
     # via boto3
 setuptools==75.8.2
     # via moto

services/dask-sidecar/requirements/constraints.txt

@@ -13,6 +13,3 @@ dask[distributed]>=2024.4.2 # issue with publish_dataset: https://github.com/das
 #
 # Compatibility/coordination
 #
-# botocore does not always add the checksums to the s3 object metadata leading to issues with CEPH S3
-# see https://github.com/ITISFoundation/osparc-simcore/issues/7585
-botocore<1.36.0

services/dask-sidecar/src/simcore_service_dask_sidecar/file_utils.py

@@ -39,7 +39,7 @@ def _file_progress_cb(
     asyncio.run_coroutine_threadsafe(
         log_publishing_cb(
             f"{text_prefix}"
-            f" {100.0 * float(value or 0)/float(size or 1):.1f}%"
+            f" {100.0 * float(value or 0) / float(size or 1):.1f}%"
             f" ({ByteSize(value).human_readable() if value else 0} / {ByteSize(size).human_readable() if size else 'NaN'})",
             logging.DEBUG,
         ),
@@ -59,6 +59,7 @@ class S3FsSettingsDict(TypedDict):
     key: str
     secret: str
     client_kwargs: ClientKWArgsDict
+    config_kwargs: dict[str, str]  # For botocore config options
 
 
 _DEFAULT_AWS_REGION: Final[str] = "us-east-1"
@@ -69,6 +70,14 @@ def _s3fs_settings_from_s3_settings(s3_settings: S3Settings) -> S3FsSettingsDict
         "key": s3_settings.S3_ACCESS_KEY,
         "secret": s3_settings.S3_SECRET_KEY,
         "client_kwargs": {},
+        "config_kwargs": {
+            # This setting tells the S3 client to only calculate checksums when explicitly required
+            # by the operation. This avoids unnecessary checksum calculations for operations that
+            # don't need them, improving performance.
+            # See: https://boto3.amazonaws.com/v1/documentation/api/latest/guide/s3.html#calculating-checksums
+            "request_checksum_calculation": "when_required",
+            "signature_version": "s3v4",
+        },
     }
     if s3_settings.S3_REGION != _DEFAULT_AWS_REGION:
         # NOTE: see https://github.com/boto/boto3/issues/125 why this is so... (sic)
@@ -96,26 +105,30 @@ async def _copy_file(
 ):
     src_storage_kwargs = src_storage_cfg or {}
     dst_storage_kwargs = dst_storage_cfg or {}
-    with fsspec.open(
-        f"{src_url}", mode="rb", **src_storage_kwargs
-    ) as src_fp, fsspec.open(f"{dst_url}", "wb", **dst_storage_kwargs) as dst_fp:
+    with (
+        fsspec.open(f"{src_url}", mode="rb", **src_storage_kwargs) as src_fp,
+        fsspec.open(f"{dst_url}", mode="wb", **dst_storage_kwargs) as dst_fp,
+    ):
         assert isinstance(src_fp, IOBase)  # nosec
         assert isinstance(dst_fp, IOBase)  # nosec
         file_size = getattr(src_fp, "size", None)
         data_read = True
         total_data_written = 0
         t = time.process_time()
         while data_read:
-            (data_read, data_written,) = await asyncio.get_event_loop().run_in_executor(
+            (
+                data_read,
+                data_written,
+            ) = await asyncio.get_event_loop().run_in_executor(
                 None, _file_chunk_streamer, src_fp, dst_fp
             )
             elapsed_time = time.process_time() - t
             total_data_written += data_written or 0
             await log_publishing_cb(
                 f"{text_prefix}"
-                f" {100.0 * float(total_data_written or 0)/float(file_size or 1):.1f}%"
+                f" {100.0 * float(total_data_written or 0) / float(file_size or 1):.1f}%"
                 f" ({ByteSize(total_data_written).human_readable() if total_data_written else 0} / {ByteSize(file_size).human_readable() if file_size else 'NaN'})"
-                f" [{ByteSize(total_data_written).to('MB')/elapsed_time:.2f} MBytes/s (avg)]",
+                f" [{ByteSize(total_data_written).to('MB') / elapsed_time:.2f} MBytes/s (avg)]",
                 logging.DEBUG,
             )
 

services/storage/requirements/_base.in

@@ -3,7 +3,6 @@
 #
 
 --constraint ../../../requirements/constraints.txt
---constraint ./constraints.txt
 
 
 --requirement ../../../packages/aws-library/requirements/_base.in

services/storage/requirements/_base.txt

@@ -2,11 +2,11 @@ aio-pika==9.5.4
     # via
     #   -r requirements/../../../packages/aws-library/requirements/../../../packages/service-library/requirements/_base.in
     #   -r requirements/../../../packages/service-library/requirements/_base.in
-aioboto3==13.3.0
+aioboto3==14.1.0
     # via
     #   -r requirements/../../../packages/aws-library/requirements/_base.in
     #   -r requirements/_base.in
-aiobotocore==2.16.0
+aiobotocore==2.21.1
     # via aioboto3
 aiocache==0.12.3
     # via
@@ -102,11 +102,10 @@ attrs==25.1.0
     #   referencing
 billiard==4.2.1
     # via celery
-boto3==1.35.81
+boto3==1.37.1
     # via aiobotocore
-botocore==1.35.81
+botocore==1.37.1
     # via
-    #   -c requirements/./constraints.txt
     #   aiobotocore
     #   boto3
     #   s3transfer
@@ -286,6 +285,7 @@ jinja2==3.1.5
     #   fastapi
 jmespath==1.0.1
     # via
+    #   aiobotocore
     #   boto3
     #   botocore
 jsonschema==4.23.0
@@ -337,6 +337,7 @@ mdurl==0.1.2
     # via markdown-it-py
 multidict==6.1.0
     # via
+    #   aiobotocore
     #   aiohttp
     #   yarl
 opentelemetry-api==1.30.0
@@ -636,6 +637,7 @@ pyinstrument==5.0.1
     #   -r requirements/../../../packages/service-library/requirements/_base.in
 python-dateutil==2.9.0.post0
     # via
+    #   aiobotocore
     #   arrow
     #   botocore
     #   celery
@@ -754,7 +756,7 @@ rpds-py==0.22.3
     # via
     #   jsonschema
     #   referencing
-s3transfer==0.10.4
+s3transfer==0.11.3
     # via boto3
 sh==2.2.1
     # via -r requirements/../../../packages/aws-library/requirements/_base.in