Cloud SQL offers excellent performance and scalability. As mentioned previously, the maximum storage capacity is 10 TB (to be distributed between HDD and SSD), while the maximum RAM that can be selected is 208 GB, and the maximum number of CPUs is 32. Google also guarantees a maximum of 25,000 IOPS, that is, input/output operations per second. The possibility to choose such resources allows us therefore to use Cloud SQL for every type of application, from those with lower workloads to applications that require intense performance. There are no limits of query per second (QPS), but there are some limits regarding the connections. In fact, up to a maximum of 4,000 concurrent connections are possible in the case of second-generation MySQL instances, while for the other cases the limits are more stringent.

For first-generation MySQL databases only, where incoming requests are briefly queuing before establishing a connection, there is a limit of 100 queued connections at the same time. The databases are also easily scalable both vertically, increasing or reducing the available resources (storage space, CPU, and RAM), and horizontally increasing the number of servers and instances that work simultaneously. Once you have created your instance, you can change the resources allocated for it directly from the console. In general, the changes are applied immediately, but the restart of the instance will close the existing connections and will need a few moments (in the case of first-generation MySQL) or a few minutes (second generation) to get back online.