Python: upload large files S3 fast

Question

I am trying to upload programmatically an very large file up to 1GB on S3. As I found that AWS S3 supports multipart upload for large files, and I found some Python code to do it. (link )

My point: the speed of upload was too slow (almost 1 min).

Is there any way to increase the performance of multipart upload. Or any good library support S3 uploading

Answer 1

Leave my answer here for ref, the performance increase twice with this code:

import boto3
from boto3.s3.transfer import TransferConfig


s3_client = boto3.client('s3')

S3_BUCKET = 'mybucket'
FILE_PATH = '/path/to/file/'
KEY_PATH = "/path/to/s3key/" 

def uploadFileS3(filename):
    config = TransferConfig(multipart_threshold=1024*25, max_concurrency=10,
                        multipart_chunksize=1024*25, use_threads=True)
    file = FILE_PATH + filename
    key = KEY_PATH + filename
    s3_client.upload_file(file, S3_BUCKET, key,
    ExtraArgs={ 'ACL': 'public-read', 'ContentType': 'video/mp4'},
    Config = config,
    Callback=ProgressPercentage(file)
    )

uploadFileS3('upload.mp4')

Special thank to @BryceH for suggestion. Although solution did increase the performance of S3 uploading, but I still open to receive any better solution. Thanks

Answer 2

1 minute for 1 GB is quite fast for that much data over the internet. You should consider S3 transfer acceleration for this use case. https://docs.aws.amazon.com/AmazonS3/latest/dev/transfer-acceleration.html