How does Boto3 S3 put_object function works in python - python-2.7

Question

With Boto3:
I am using put_object() function to upload object in s3. I am using put_object() with customer encryption key parameter for server side encryption.
With Boto:
I am using upload_chunk function to upload object in s3. Here I am using aws managed keys for server side encryption and not customer given as it is not supported in API
So With Boto3 approach my program is taking more memory than Boto approach.
Please tell me how does put_object function works in boto3 for server side encryption .
is it using memory of machine for encryption on which it has called ?
should I explicitly clean data buffer which is passed as Body parameter to put_object function ?
Code:
def put_s3_object(self, target_key_name, data, sse_cust_key, sse_cust_key_md5):
''' description: Upload file as s3 object using SSE with customer key
It will store s3 object in encrypted format
input:
target_key_name (#string) data(in memory string/bytes)
sse_cust_key (#string)
sse_cust_key_md5 (#string)
output: response
'''
if not target_key_name:
raise
try:
response = self.s3_client.put_object(Bucket = self.source_bucket, Body = data, Key = target_key_name, SSECustomerAlgorithm = awsParams.CLOUD_DR_AWS_SSE_ALGO, SSECustomerKey = sse_cust_key, SSECustomerKeyMD5 = sse_cust_key_md5)
del data
except botocore.exceptions.ClientError, fault:
raise
except Exception, fault:
raise

Answer 1

Instead of Boto3 put_object(), We can go with set_contents_to_string() and get_contents_as_string() functions from boto.
These calls also supports server side encryption with customer keys(SSE-C). we just need to give all keys information in headers
For more details
http://docs.aws.amazon.com/AmazonS3/latest/dev/ServerSideEncryptionCustomerKeys.html