Azure SQL Elastic Pool max size lower than max size of database included in pool

Question

I am trying to save some money by utilizing Azure SQL Elastic Pools. I have a bunch of databases which only peak in usage once a while, so having them in a pool would be optimal. The problem now is that all the databases combined need about 1 TB of storage, which I only get by selecting the standard service tier and selecting 300 DTUs. After trying out some things I noticed that even though the size of the elastic pool is capped to a value (e.g. 500GB with 50 DTUs Standard), I can increase the size of an included database to above the cap (e.g. 750GB, even though the pool size is only 500GB).

My questions now are:

• What exactly happens if I increase the database size over the pool size? Am I going to get billed for extra GB as soon as I go over the max pool size?

• More a side question, but is there another option to increase the storage size without raising the DTUs so much? I would need maybe like 100 DTUs with 1TB of storage. I looked into the vCPU pricing options, but they seem even worse pricing wise.

Answer 1

Increasing the size of a database within an Azure SQL Elastic Pool beyond the pool size limit is allowed, but it's important to understand the implications:

1. Billing: Yes, if you increase the size of an included database beyond the pool size limit, you will be billed for the additional storage space used by that specific database. The storage cost for that database will be calculated separately from the pool.

2. Performance: While you can increase the database size, the performance (DTUs) remains tied to the pool. So, if your pool has 50 DTUs, all the databases in the pool will share those DTUs. Increasing an individual database's size won't increase its performance unless you adjust the database's DTU settings separately.

Regarding your side question, if you need more storage (1TB) without significantly increasing DTUs, you might want to consider Azure SQL Database Hyperscale. Azure SQL Hyperscale offers on-demand storage, so you can independently scale storage without affecting the DTU cost. You can start with lower DTUs and allocate the storage you need. Keep in mind that Hyperscale pricing is different from the standard Azure SQL Database pricing, so you should review the pricing details to ensure it meets your budget and requirements.