If you are new to cloud architecture this old book (first published Sept 2012) can be your primer. It was written for those who want to get involved in cloud architecture conversations. This book gives the basic understanding of cloud patterns such that you can leverage them in architecting your cloud-native applications. To miximise the value of cloud, cloud-native applications, like Netflix, Twitter, LinkedIn etc are built to leverage the cloud patterns. The patterns in this context are engineering practices that have proven successfully among cloud service providers like Microsoft Azure, Amazon Web Services (AWS) and Google Cloud. Although the examples in this book refer to Microsoft Azure, I found the concepts and techniques correlate with other cloud platforms.
The essence of this book is the patterns, the challenges you are likely to encounter in the cloud and architectural solutions using the patterns. As usual, there is always something new to learn, so here are three architecture patterns that sparked my interest;
- Queue-Centric Workflow Pattern: This cloud pattern decouple communications between web tier (which implements the user interface) and service tier (where business processing happens). A service tier which is unreliable or slow can render a bad experince to a user on web tier. The solution is to allow sender and receiver to communicate only through messages on a queue. This is how Twitter, LinkedIn, Facebook etc handle your command to create new user account, add photo or update status.
- MapReduce Pattern: MapReduce is a data processing approach focusing on big data using a parallel and distributed algorithm on a cluster. Almost all cloud platforms offer MapReduce as a service. Therefore LinkedIn (on Microsoft Azure) uses it to suggest contacts you might want to add to your network, Amazon (on AWS) uses it to recommend books and Twitter (on Google Cloud) uses it to recommend Twitter’s accounts.
- Database Sharding Pattern: This pattern focuses on horizontal scaling of data in a database such that each shard is held on a separate database instance and each shard has the same database schema as the original database. Therefore the combined data from all shards is the same as the data from the original database. All cloud platforms offer integrated database sharding as a database service, in both relational and NoSQL varieties.
Again these patterns in addition with other patterns in this book were covered at very high level for beginner kind of audience or manager type of folks therefore if you are an architect you would definitely want some more stuff and therefore I recommend you supplement it with other books in this same category.
