as long as you have an AWS account, basic SQL knowledge, and know how to modify IAM policies you can do everything in this book. We tried to really make it accessible to beginners and intermediate folks alike. Only one or two chapters at the end are a bit more advanced and do require comfort with developing java code.

In this case we typically mean that you do not need to run jobs/queries that read literally ALL your data and then rearrange it or pre-join it with other data to make it either (a) small enough (b) in the right format or (c) denormalized across data sources. These techniques can still be helpful with Athena, but Athena does its best to still make data that is no well organized, optimized, or formated using best practices accessible without all the upfront work/maintenance of ETL

Yes, we typically mention it in the chapters but while authoring the book we spent < $30 total on AWS fees and keep in mind we had to run through the exercises 5 or 6 times to refine them.

Aside from the visualizations that you get with Tableu I think youll find the experience with Athena is pretty similar with one key difference being that if you ever need to scale up from 100k records to 100Billion or somewhere in between Athena will handle that just as well and you might start to see other differentiation in terms of price and performance as you scale up. But id agree that at 100k records, probably even Excel can used for a bunch of quick analysis.

We tried to include practical examples in every chapter in the form of exercises. There are also lots of stories about cases we’ve seen people use Athena for X or Y. As for AWS knowledge, the pre-requiste is pretty small, just that you (a) have an AWS Account (b) know some basic SQL (c) are able to modify IAM policies in your AWS account since each chapter requires different levels of access (though you could just do the entire thing with admin access if the account your using isn’t a production account or is a ‘burner’ as we say).

There is a chapter in the book with some basic examples on how to create what many would call an ETL pipeline but if your needs include chains of jobs (probably anything more than 3 or 4 jobs). I’d recommend looking at Glue ETL as it has the concept of scheduled jobs as well as a dependency modeler that can run chains of jobs based on triggers like data arriving in S3, completion of another job, or a schedule.

Thanks!

what do you mean by who?

ah sorry, I meant how, sometimes my autocorrect behaves strangely

Redshift targets datawharehouse usecases while Athena is more geared towards ad hoc analysis without ETL. Keep in mind this is an extremely reductive explaination and the two products can both do many of the same things just with different price/performance and strengths.

it won’t replace Redshift + Spectrum as those services do a lot more than just ACID. They have a different performance profile, cost, and SQL feature set.

I do think youll see more overlap as time goes on since AWS is striving to make it easy for customers to move between and/or use a combination of services as one seemless offering since each it fit for a specific purpose. For example, can you use a hammer to break concrete? Yes, but a sledge hammer would be btter. and You can use a sledgehammer to drive nails but a hammer would be better.