Book Image

Google Cloud Platform for Architects

By : Vitthal Srinivasan, Loonycorn , Judy Raj
Book Image

Google Cloud Platform for Architects

By: Vitthal Srinivasan, Loonycorn , Judy Raj

Overview of this book

Using a public cloud platform was considered risky a decade ago, and unconventional even just a few years ago. Today, however, use of the public cloud is completely mainstream - the norm, rather than the exception. Several leading technology firms, including Google, have built sophisticated cloud platforms, and are locked in a fierce competition for market share. The main goal of this book is to enable you to get the best out of the GCP, and to use it with confidence and competence. You will learn why cloud architectures take the forms that they do, and this will help you become a skilled high-level cloud architect. You will also learn how individual cloud services are configured and used, so that you are never intimidated at having to build it yourself. You will also learn the right way and the right situation in which to use the important GCP services. By the end of this book, you will be able to make the most out of Google Cloud Platform design.
Table of Contents (19 chapters)
13
Logging and Monitoring

Use BigQuery unless you have a specific reason not to

There is an absolute candy store of data storage services available on the GCP. You probably have seen a taxonomy like the one in the following diagram, which tells you which service to use when. Now you certainly should take the time to go through this diagram and understand each part of it, but the bottom line can be summed up as: unless you really have a strong pressing reason, just use BigQuery:

One asterisk about this preceding taxonomy is we intentionally list Datastore in the OLAP part of the tree because even though it does have a transactional mode, in the real world, if you need transactional support, you are far more likely to go with RDBMS than with a document-oriented NoSQL database such as Datastore.

What might some of those strong pressing reasons to not use BigQuery be? Well, you might need really ironclad...