Wednesday, January 9, 2013

Database and disk size estimation

Since I am getting a lot of questions regarding estimation of database sizes, I have calculated that spending too much time to predict exact size of database is much more expensive that spend money to larger disk.
Many customers want to have exact size, and formula how we estimated database size. That estimation is directly connected with money, since when we give an estimation, customer is going to buy disks of such size.
So, in order to calculate how much time is profitable spending on this task, we need only two numbers - SAN price per GB(with all costs) and how much are worth hours of people who are included in process of database (and disk) size estimation.
Very rough estimation for SAN storage is varying between several dollars to maximum 40 dollars per GB.
Also, if You also include in calculation well known Pareto rule which said that with 20% of time You will reach 80% of You goal, it means that we do not have to spend too much time on disk size prediction - just tell our customer to buy big enough disk, and that if there is no enough space that we will need to buy more...If any of customers are willing to accept this explanation.. :-)