I'm setting up some infrastructure for a small playground website and want to keep the database as free or inexpensive as possible while I'm developing. Azure's free offerings appear to include quotas for certain databases, but I keep finding that there's no available quota in the regions I've tried. I checked Cosmos DB, Azure SQL Database, and MySQL across UK South, West Europe, North Europe, East US, and East US 2. Is there a better way to determine which database services and regions actually have free capacity? I'd also appreciate recommendations for a cheap relational database suitable for a lightweight project.
3 Answers
For a small workload, serverless Cosmos DB is worth considering. It can cost almost nothing at low usage, although you should design your application to handle 429 throttling responses. Storage account table storage is another very inexpensive option, but it’s much more limited than a relational database. Cosmos DB generally has capacity in many regions and usually doesn’t require the same quota request process, so it may be worth checking that your subscription has the provider registered and trying a few regions.
Azure SQL Database is probably the simplest relational fallback. The Basic tier is around $5 per month, while a small DTU-based tier may be closer to $15 monthly. For a lightweight development project, Basic or a low DTU configuration may be sufficient. The Azure pricing calculator can help compare the purchase model and DTU options before you create the resource.
That price range will probably be my fallback. I originally thought the database options would be fully free, but several services don’t seem to have remaining quota in the regions I checked.
Azure Table Storage is extremely cheap and can work well for simple key-value or document-style data. It isn’t a replacement for SQL if you need joins, complex queries, or relational constraints, but it’s a good low-cost choice for straightforward data access.
That sounds promising—thanks. I’ll look into provisioning Cosmos DB and handling the throttling behavior.