The serverless conversation is always focused on compute. This means that the serverless database is usually left until the last moment.
It’s worth spending some time upfront to consider what you need from your serverless data store. And to look at what the main cloud providers and vendors can offer.
The 7 tenets of serverless data
The 7 tenets listed are:
- No operational overhead – No heavy lifting
- Elastic – all-consuming
- Reliability – Always available
- Performance – predictable and consistent
- Universally consistent – always correct
- Security – web-native access
- Cost-optimized – usage-based pricing
Ultimately, the tenets help you select a data solution that will integrate well with your serverless architecture. Have a read of the main article. Similar to compute, serverless data is an entirely different paradigm. Selecting a serverless database is different, so there’s a different headspace you need to get into for a few things. One thing is for sure; it’s a powerful and exciting approach and one that will continue to rise in popularity.
Are all databases serverless?
We have previously covered some fundamental aspects of becoming a serverless-first engineer, but this article goes beyond that. It’s not advisable to create your own data solution. You have to use a service from your cloud provider or a vendor. The tenets are a guide to which databases are serverless. Like anything popular, many services declare themselves to be serverless because they have a pay-per-use model. Paying your bill is one thing, but your data needs to scale, be reliable and perform – with no operational burden. Imagine your application has a huge spike, and your database falls over? This shouldn’t happen with a genuine serverless database. The tenets will help you decide. For the record, I think Fauna is an excellent choice and well worth a look, but go and decide for yourself. Read the tenets, figure out your requirements and try a few out.