Amazon S3 自定义指标报告

Question

I have a requirement where large number of data package files (example:- xml,json,jpeg files) will be send to Amazon S3. How can we get the aggregate count of data transmitted based on some input data points(for example- timestamp). Also how can I also get the information of the read/write time taken by Amazon S3. I have heard about Amazon Cloudwatch for custom metrics but I have no idea how to implement this use case using Amazon Cloudwatch.

Answer 1

Basic approach is described in documentation . Two default metric exists in every bucket, first to get files count and second to get files size. How to get size of files is in doc and to get count of objects in bucket with name: 'ExampleBucket' use aws cli with this command:

aws cloudwatch get-metric-statistics --metric-name NumberOfObjects \
  --namespace AWS/S3 --start-time 2020-10-19T00:00:00Z \
  --end-time 2020-10-20T00:00:00Z --statistics Maximum --unit Count \
  --region us-west-1 \
  --dimensions Name=BucketName,Value=ExampleBucket Name=StorageType,Value=AllStorageTypes \
  --period 86400 --output json

For more code examples see this question .