I have a bunch of s3 folders for different projects/clients and I would like to estimate total size (so I can for instance consider reducing sizes/cost). What is a good way to determine this?
Asked
Active
Viewed 2,610 times
1
-
Similar to: [How to find size of a folder inside an S3 bucket?](https://stackoverflow.com/q/49759940/174777) – John Rotenstein Mar 30 '20 at 03:33
-
Sub-folders too, or just top-level folders? – John Rotenstein Mar 30 '20 at 03:34
4 Answers
1
I can do this with a combination of Python and the AWS client:
import os
bucket_rows = os.popen('aws s3 ls').split(chr(10))
sizes = dict()
for bucket in bucket_rows:
buck = bucket.split(' ')[-1] # the full row contains additional information
cmd = f"aws s3 ls --summarize --human-readable --recursive s3://{buck}/ | grep 'Total'"
sizes[buck] = os.popen(cmd).read()

Sergio Lucero
- 862
- 1
- 12
- 21
-
`aws s3 ls s3://bucket/path/to/data/ --summarize --human-readable --recursive | grep 'Total'` is an excellent way to quickly get subtotal size. Thanks! – Wassadamo Nov 30 '21 at 07:56
1
As stated here AWS CLI natively supports --query parameter with can determine the size of every object in S3 bucket.
aws s3api list-objects --bucket BUCKETNAME --output json --query "[sum(Contents[].Size), length(Contents[])]"
I hope it helps.

Vaibhav Jain
- 2,155
- 5
- 27
- 41
1
This would do the magic
for bucket_name in `aws s3 ls | awk '{print $3}'`; do
echo "$bucket_name"
aws s3 ls s3://$bucket_name --recursive --summarize | tail -n2
done

Jorge Tovar
- 1,374
- 12
- 17
0
If you want to check via console.
- If you mention a folder by folder and not bucket, then just select that object go to "action" drop down and select "Get total size"
- If you mean bucket by folder , then go to management tab and in that go to metrics it will show entire bucket size

Abhishek Kumar
- 143
- 1
- 7