Reputation: 29
I am confused about pricking criteria in aws s3.
Let's assume
SELECT * FROM TABLE => result 1000 rows
Is this command 1000 requests or 1 request?
Upvotes: 1
Views: 392
Reputation: 128
AWS S3 Select charges for the bytes scanned ($0.00200 per GB) and the bytes returned ($0.0007 per GB).
The charges are broken in your billing management console Under Billing>Bills>Details>Simple Storage Service>Region> "Amazon Simple Storage Service USE2-Select-Returned-Bytes" and "Amazon Simple Storage Service USE2-Select-Scanned-Bytes"
Using limit, where clauses or having a compressed file reduces the bytes scanned. The service supports gzipped Parquet files which is fantastic. I was able to reduce file sizes by 90% and take advantage of columnar data format rather than csv.
Upvotes: 0
Reputation: 6333
If you using S3 Select to query, it is a single request and you get charged for every 1000 requests and charged at the same rate as object GET requests
If you are using Athena to query S3, the charge would be for the amount of data retrieved and how the file is stored like zip or parquet format
Upvotes: 2