Transaction Management Techniques of Concurrency and Recovery

2847 Words Feb 21st, 2018 11 Pages
So choosing the right database to fit the needs of the business at the costs available has become an important job. What this paper aims to do is discuss the ideals behind relational databases and NoSQL databases as well as compare both “open source” to premium services focusing on transaction management techniques of concurrency and recovery of said database. That was a mouthful. To say it in a simpler way, “Which type of database best fits the business?” and “What do I get for my money?”

Relational databases operate under a set of properties that are abbreviated as ACID (Atomicity, Consistency, Isolation, Durability). Every individual, logical operation in a database is called a transaction. If you read, update, delete, etc.. any field in a table it is said you have performed a transaction. In a relational model the transaction should be “all or nothing”, which is what is meant by Atomicity. After a transaction the database should be valid as laid out by the specified schema. This is known as keeping the database in a consistent state. Concurrency control aims to make sure every transaction performs the task it was meant to do with the correct results being met or else “rolled back” or reset to where the database was before the transaction…
Open Document