Monthly DynamoDB procedure - best practices help

Question

We run a system that collects data monthly, so we have several central tables, then more of a YYYY_MM structure.   So days before the new month we are creating a table, setting capacity, PITR, etc.  Then around the 6th, we will lower the capacity from the previous month, lower capacity, etc. and I want to automate this.  The only thought I had was using an admin server with AWS keys, and writing a script using CLI tools.   Naturally that user would need credentials, etc. but is there any other way that would be easier?

Also regarding the backup, if I am using AWS Backup, once the previous month is done, I am trying to find out how I am charged, if the data doesn't change and I leave the daily flag on, do I pay for more storage (the same data over and over) or is it just the one time as the data is the same as that will affect that part.

Thank you in advance.

Accepted Answer

My suggestion here would be to use [Amazon EventBridge](https://aws.amazon.com/eventbridge/) which invokes a Lambda function on monthly basis. The logic in your Lambda can be used to change the tables capacity settings etc.. I would also suggest looking into [DynamoDB Standard-IA](https://aws.amazon.com/dynamodb/standard-ia/) table class which may suit your use-case.

For Auth, you can attach an IAM role with the necessary permission to your Lambda function so it can make the desired updates. More [here](https://docs.aws.amazon.com/lambda/latest/dg/lambda-permissions.html)

For DynamoDB Backups, you are only charged for the [data stored](https://aws.amazon.com/dynamodb/pricing/on-demand/), so if the data does not change then your monthly bill will remain the same.

Answer

Thanks Leeroy, have never used EventBridge due to I am a bash/linux guy and Lambda doesn't support, but I keep saying I need to get on to Python so here is yet another reason!

Thanks much

Monthly DynamoDB procedure - best practices help

相關內容