Seven Databases in Seven Weeks

I just finished this book and I really enjoyed it.Let me mention just for context, that I've been a Data Scientist for 10 years now and have been working with No-SQL databases since CouchDB - whereas I use CosmosDB and Cassandra day to day. I've been using NOSQL for all of that time, and have spent a lot of time transitioning. Let me go through a few observations and I'll connect the dots afterward.The biggest thing you'll get from this book is a reinforced understanding of where NOSQL is ideal for solving a problemset and where it's not appropriate. You can make anything work, but each NOSQL Db has specific features aimed at solving certain problems. Cassandra may do X, Y & Z at some huge scale but if you don't have X, Y or Z as needs or goals, if you don't need those features, it will likely not just be a waste, but you could end up adding unwanted complexity that has huge legacy costs. There are many problems NOSQL databases as a group solve well. In that space certain ones might handle things better than others. But you should not move to NoSQL or adopt it just b/c it's what everyone is doing. SQL databases have been around a long time and have proven themselves to be very solid in many different domains. Understanding where and why gives you so much insight into what to use that you really want to make sure you have a firm grasp on this before any serious undertaking.This is a later release of the Seven Databases in Seven weeks original version which was a great way to get acquainted with the material. It's impossible to become truly skilled at any enterprise grade database in a week. However if you go through each chapter diligently and experiment on your own, make a point of looking up things you don't know and drilling areas you are weak in (even looking up areas that you don't feel strong in), after a week you'll find you probably know enough to get through most job interviews on the subject. I'd add that if you read the previous version you'll have a big head start, rather, you'll learn a lot more b/c the ideas are already in your head. The same is the case here, you'll get more out of the later chapters than the early ones, just b/c the early ones serve as a basis (not b/c they are not as good or as useful.) The book transitions in a very intuitive manner.It starts out with PostgreSQL. It covers the basic features (CRUD, Queries, Indices, Rules) and then wraps things up with using Full-Text and Multidimensional features. Multidimensional is usually separated out for instance, in SQL Server it's called Analysis Services and is Microsoft's solution for OLAP warehouses. Postgres is a great product and you can tell that by the sheer love most of the user community has for it. Note that they covered both Full-Text indexing and MDM, I suspect not just b/c they are advanced features but things that Document Databases help with although in a totally different way.Next is HBase which IMHO, you really need to at least be familiar with if you're dealing with either large sizes or high velocity data. MongoDB and CouchDB are covered next, Document databases. I've interviewed people that supposedly knew both and took a position that Mongo was just an improved CouchDB. They weren't aware of the big differences with features like Sharding, GeoSpatial data storage, GridFS or Replica Sets. If you want to use Mongo, this sets you up for a really good experience. You'll want to take either chapter or both and do a little extra homework but this book guides you there perfectly.NEO4J is covered next with an emphasis on REST, Indexes Algorithms, and Distributed High-Availability. The discussion on DHA alone warrants the price of the book - I hadn't used NEO4J before this but I gained a real appreciation of it.Next is DynamoDB which i hadn't used before and I've actually read it three times - not b/c I didn't understand it but b/c it gives you profound insight into what set of problems it's geared toward. At this point, we're getting a lot of feature overlap so understanding the distinctions becomes critical.Lastly is Redis. It begins with CRUD and DataTypes, moves on to Distrubtion and ends with interaction with other databases. This chapter was pure gold as well. Getting these insights in a condensed and easy to understand form is excellent.At the end, you'll have a good sampling of databases in the NoSQL Space. The way they chose the sample, you'll also come away with a great understanding of the features provided. So, they may not have discussed some of the higher profile document databases, key/value stores or grid dbs, but you'll understand features and when you look at other systems that highlight those features, you'll have a good understanding of them from the start.I was particularly impressed with the discussion on NEOJ4 and the last two but for '7 days' they walk you through so much, never wasting anything on features you won't use often. I really appreciate how concise and efficiently they covered all of this - not wasting your time on features you won't likely use or care about unless you have some very rare scenarios while giving you a really good foundational understanding of each.Look, if you're building a new app or company, getting the database right is a big deal. Moving data stores is a hassle that many people take too lightly only to learn the hard way that they should have been more thorough. Knowing what to avoid, what mistakes people make, the problem set that these items are suited to and what they aren't suited to, all the while making sure you know what is there for you now, makes this book another winner every bit as good as the last edition, if not better.Even if you're just reading to keep up to date, this book is one I think pretty anyone should have.

While I use Microsoft's SQL Server for commercial development, it's always good to keep an eye on other technology stacks. In particular the so-call NoSQL technologies. Now everything is connected, or related, but it's good to have a quick go-to database that simply holds documents. This book is not limited to NoSQL, it talks about PostgreSQL, nut also big unrelated data repository like HBase. The book also talks about graphing data, like maps, using Neo4j, whihc optimizes path finding.I was most interested in learning about CouchDB. It falls in the big unrelated data category. It's under the Apache umbrella, so you know it's good. But there's another reason why CouchDB is particularly robust, it's written in the Erlang, , which is a functional programming language, which uses immutable resources, which means they're not shared, which means there are no race conditions, which means you have a solid distributed piece of software that doesn't crash. Perfect for a web application don't you think?CouchDB is readily available in the Internet to be downloaded, and easy to install. Once it's running, it's a pleasure to see the JSON output if you visit the localhost at port 5984:{"couchdb":"Welcome","version":"2.1.1","features":["scheduler"],"vendor":{"name":"The Apache Software Foundation"}}Everything in CouchDB is handled through RESTful services.With a NoSQL database, like CouchDB, multi-core parallelism is really important for efficiency in using MapReduce algorithms over massive data. As I mentioned earlier, Erlang is really, really robust in multi-threaded development.

The figures in the graph DB section are faded and completely unreadable. One would think that the copy would be proofed before people pay good money for it. The reasons for using certain database types are not entirely convincing.Otherwise it is a decent introduction to the various database types.

Honestly, I borrowed this book and read maybe 20%, but I plan to finish if allowed by time. The book does not have installation tutorials on any of these NoSQL systems, which in half of the cases are tedious and depend on your operative system.It discusses their strengths, basic and intermediate query notations, the languages that they rely on and some scripts to enhance their capabilities, the basic background of what's underneath, what is required to start using it.Each database is cover in a chapter, each one with an introduction, 3 subsections to be covered in 3 days, and HW suggestions, which you must do to progress to the next day.

I've read a number of the "7 X's in Seven Weeks" books, and this is clearly the best so far. Other books have just focussed on new and emerging technologies, with the implication that "you're going to have to change to something new, so let's look at the alternatives". Well, no. Most innovation is accretive, adding new tools to your repertoire, and the fact is that sometimes the older, well-established choice is the best for the job. So "Seven databases" starts off with the venerable PostgreSQL (which evolved from the original INGRES), and builds on this. The result is a book that not only provides a balanced survey, but that also allows the reader to "compare and contrast" in order to decide if and how to adapt their practices.Thoroughly recommended.

Not sure which database you need? This book gives you an awesome primer on each one that includes getting setup, differentiating features, and some hands on examples. I highly recommend this to anyone that wants a primer on the various databases that are out there.

Databases enlisted are chosen by the author in a very thoughtful way. Every presents certain traits, which are well documented by the book. It’s not only about how to use other databases, but also which and when to choose. Warmly recommend!

Ottima panoramica di approcci tradizionali ed innovativi al DB, con esempi concreti e ben guidati rende operativi sui 7 prodotti indicati e fornisce linee di approfondimento.