Align PutObject checksum error messages with S3

[cbornet]

Expected Behavior

• Should return The CRC32 you specified did not match the calculated checksum. when a PutObject fails due to a CRC32 checksum mismatch.
• Should return The CRC32C you specified did not match the calculated checksum. when a PutObject fails due to a CRC32C checksum mismatch.

Current Behavior

• Returns The provided 'x-amz-checksum' header does not match what was computed. when a PutObject fails due to a checksum mismatch.

Possible Solution

Steps to Reproduce (for bugs)

1. Launch a local silo server
2. Run the script

import boto3
import zlib
import base64
import tempfile
import os
from botocore.exceptions import ClientError

os.environ["AWS_REQUEST_CHECKSUM_CALCULATION"] = "when_required"
CHUNK_SIZE = 1 * 1024 * 1024  # 8 MB

AWS_REGION = "us-east-1"
BUCKET_NAME = "your-unique-bucket-name"


def crc32_to_base64(crc):
    crc &= 0xFFFFFFFF
    return base64.b64encode(crc.to_bytes(4, "big")).decode("utf-8")


def create_s3_client():
    return boto3.client(
        "s3",
        endpoint_url="http://localhost:9000",
        aws_access_key_id="minioadmin",
        aws_secret_access_key="minioadmin",
        region_name=AWS_REGION
    )


def create_bucket_if_not_exists(s3, bucket_name):
    try:
        s3.head_bucket(Bucket=bucket_name)
        print(f"Bucket already exists: {bucket_name}")
    except ClientError:
        print(f"Creating bucket: {bucket_name}")

        if AWS_REGION == "us-east-1":
            s3.create_bucket(Bucket=bucket_name)
        else:
            s3.create_bucket(
                Bucket=bucket_name,
                CreateBucketConfiguration={
                    "LocationConstraint": AWS_REGION
                }
            )


def create_temp_file():
    tmp = tempfile.NamedTemporaryFile(delete=False)
    target_size = CHUNK_SIZE * 1  # 3 parts
    chunk = b"A" * (1024 * 1024)  # write in 1 MB blocks

    written = 0
    while written < target_size:
        to_write = min(len(chunk), target_size - written)
        tmp.write(chunk[:to_write])
        written += to_write
    tmp.close()
    return tmp.name


def upload(s3, file_path, bucket, key):
    with open(file_path, "rb") as f:
        data = f.read()

    crc32 = zlib.crc32(data) + 1
    crc32_b64 = crc32_to_base64(crc32)

    result = s3.put_object(
        Bucket=bucket,
        Key=key,
        Body=data,
        ChecksumCRC32=crc32_b64,
    )

    print("Upload complete ✅")
    print("S3 checksum:", result.get("ChecksumCRC32"))


if __name__ == "__main__":
    s3 = create_s3_client()

    # 1. Ensure bucket exists
    create_bucket_if_not_exists(s3, BUCKET_NAME)

    # 2. Create temp file
    temp_file = create_temp_file()

    try:
        # 3. Upload
        upload(
            s3,
            file_path=temp_file,
            bucket=BUCKET_NAME,
            key="uploads/temp-example.txt"
        )
    finally:
        os.remove(temp_file)
        print("Temp file deleted")

Context

Aligning Silo with S3

Regression

No

Your Environment

• Version used (silo --version): RELEASE.2026-09-16T00-00-00Z
• Server setup and configuration:
• Operating System and version (uname -a):