NoSQL database for the banking system

Question

Is this a good decision to use the NoSQL database for the banking system, and not for the RDBMS?

If so, what are the recommended NoSQL databases for the banking system?

Answer 1

Nathan Hurst has a really good blog post about the idea of ​​NoSQL databases . I will do my best to rephrase:

The database is usually selected based on the properties of consistency, accessibility, and partition tolerance ( CAP Theor ). Of course, the CAP theorem states that a database can really focus on only two of them. NoSQL databases require scale tolerance to scale properly, so they end up sacrificing either accessibility or consistency. RDBMSs discuss this issue by choosing consistency and accessibility and using other means to maintain the tolerance of the data section (for example: replication).

You can usually see the effects of this at the transaction level. In a DBMS land, all transactions must be ACID (Atomic, Consistent, Isolated and Durable). NoSQL databases usually do not have strict ACID requirements. Thus, data that is updated through a transaction may or may not be atomic (the transaction is either completed to all places of updating, or rollback), may be short-lived if a power failure occurs and can be performed in accordance with the assumption of a "possible consistency".

So no, the NoSQL database is definitely not a good idea for a banking solution.

It should also be noted that the NoSQL database architecture is significantly different brand. What I said here is a generalization of NoSQL databases. This, of course, is not all-encomapssing.

Answer 2

Before answering this question, I would like to give an example: GT.M is a NoSQL database that provides extreme transaction. which is used in the world's largest banking system, FIS Core Banking System (rating # 1 from inntron )

Thus, it is theoretically possible to use NoSQL for major banking systems, provided that your NoSQL engine supports transactions.

Source: http://www.slideshare.net/fachrybafadal/nosql-technology

Answer 3

Having worked in the banking industry, I would be careful in implementing anything other than proven, "outdated" RDBMS systems or mainframes for basic banking purposes (accounts, SOR, GL, etc.). For peripheral systems such as marketing, analytic databases, etc., NoSQL is fine, but you need to be more specific in your use case to get any good answer to this question.

Each tool has the right use case. Banking is extremely risky and conservative.