Database Migration: How Long Should it Take?
Database migration takes time. But how long, exactly, should you expect a database migration to last? Which factors influence the length of time that data migration requires? We explore in this article.
What is database migration?
A database migration is any type of movement of data from one database to another.
Database migrations come in many forms. They could involve simply moving data from one database instance to another database instance of the same type; for example, you might transfer data on a MySQL database located on one server to another MySQL database on a different server.
Data migration could also entail multiple types of databases. You could move data from MySQL to Cassandra, for instance. In that case, you may need to perform some data transformations.
In still other situations, a database migration could mean moving data from one type of storage medium, like tapes, into another, like hard disks. This type of database migration is common when moving data between backup and production systems.
The data conduits for database migrations can also come in multiple forms. Transferring data via a network connection is the most common approach these days, but you could also move data from one disk to another within the same system, or use portable disks (like external hard drives or DVDs) to transfer data between servers.
Read our eBook
See how four companies are using Connect for easy, automated, real-time database replication to drive their businesses forward
How long does database migration take?
Asking how long database migration takes is like asking how much it costs to go on vacation or how many bedrooms you need in a house. The answer will vary widely depending on a variety of factors.
That said, when planning a database migration, it is possible to get a sense of which factors will influence migration time. The three chief factors include:
- How much data you are transferring. This is the most obvious, and usually the biggest variable in shaping transfer time. More data means a longer migration time.
- How you are moving the data. Will you transfer data via the network? If so, how much bandwidth does your connection support? If you are copying directly from one disk to another, what is the read and write I/O of the disks? (Keep in mind that theoretical speeds are often difficult to achieve in practice in these situations.)
- How much data transformation is required, and how streamlined your process is for performing data transformation. The more transformation you have to do, and the more you rely on a manual process for transforming data, the slower your database migration will be. This is especially true in situations where you have to move data from one type of database to another.
To optimize database migration speed, focus on controlling these factors.
Data transformation and database migration speed
Before we wrap up, let’s revisit that third factor on the list above: Data transformation. In many ways, data transformation is the most important factor to focus on if you want to improve database migration speeds.
The reason is that data transformation is the only factor over which you have a lot of control in most cases. You typically can’t control how much data you transfer; you have to work with whatever the size of your databases already is. Also, you can’t often control how you transfer the data; you have to use the transport mechanisms available to you.
But what you can control is how your data is transformed, and how automated the process is. Ideally, you’ll take advantage of data transformation tools that allow data to move seamlessly from one type of database to another, with no manual transformation work required on the part of human operators. This is one area where Precisely’s data integration solutions can come in handy for certain types of database migrations.
For more information, read our eBook: Customer Use Cases for Real-Time Database Replication