I will tell you Oracle DataGuard technology and architecture that we use to protect our databases from possible disaster scenarios.
Before begin to describe the Oracle Data Guard, let us state what the Disaster states are for our databases. When the disaster in daily life is mentioned, such as the earthquake, floods, explosions, tsunami, many goods and events that cause loss of life are known.
Lets think about these disaster events for companies, one of the most important disaster events for companies is the loss of the data in the Databases that contain critical data for any reason. So It is very very important to protect enterprise data because they store customers’ data in their databases like banks, holdings, telecommunication.
If we example in the daily life, you are making money transactions with x bank. What happens if all databases of the X bank fail or corrupted and do not have access to customer data in any way? Then X Bank doesn’t know How much money, gold, dollar for every customer in bank X, if there is no data. This scenario I mentioned is a disaster situation for banking and for many other companies. When companies are exposed to such disasters, they must use Disaster Recovery solutions known as Disaster Recovery, which their databases offer to get rid of disaster.
Companies must find answers to the following 2 questions before implementing Disaster Recovery solutions and make sure that these answers meet the needs of the company.
1. How long can we work properly without relevant business critical data?
2. How much data we will lose in case of disaster?
For the database of a banking application, the answer to the above questions will be as short as possible and not to lose any data at all.
Each database management system (Oracle, SQL Server, MySQL, Sybase, etc.) provides any solution against disaster situations that I mentioned above. Oracle Corporation is released disaster solution at the first Oracle 7 to provide a manual standby database as a Disaster Recovery solution. Oracle has introduced the disaster recovery solution with Oracle 8i as the Data Guard, and many features have been added to almost every version.
With Oracle Data Guard, we can create a Backup Database which is replicated by live data of our database very quickly. Thus, we can proceed from the Backup database very quickly at the moment of any disaster in the main database. Here, one of the most important roles of Dataguard is that the master database not only creates a Backup database but also creates an instant backup based on the configuration created because it feeds backup with live data.
The Architecture of Data Guard can be summarized as follows. Dataguard moves the Redo records generated in the Primary database, which we call the Live database, to the Standby database, which we call Backup database or disaster database, and this Redo data is applied to the Standby database and creates most Current backup of the Primary database everytime.
The data that we call Redo Data is a complete set of Transactions generating in the Primary database (Inset, Update, Delete, Create, Alter, etc.). Moving the transactions from the Primary database to Standby means applying the same transactions to the Standby side. The Data Guard also checks the Redo data it carries before it applies to the Standby database, but also against Block Corruption problems.
Data Guard has 2 different modes of operation: Synchronous and Asynchronous.
I will continue to tell these modes in the next post.