• 2 min read

DocumentDB: Everything you need to know about using the new pricing options

Last week at //Build 2016, we made some big DocumentDB announcements, including changes to the pricing and scaling options. Read on to learn more.

Last week at //Build 2016, we made some big DocumentDB announcements, including changes to the pricing and scaling options. As part of these changes, DocumentDB decoupled storage size from throughput in a new pricing option, enabling pay as you go storage and a user-defined throughput level that’s customizable for your application needs.

Already using DocumentDB?

If the size of your existing 10GB single-partition collection suits your needs, and you simply want to take advantage of the new user-defined throughput levels (up to 10,000 Request Units per second on single-partition collections), you can make the change in the Azure portal with just a couple of clicks. For instructions, see Changing performance levels using the Azure Portal, or watch the 75-second video on Channel 9.

When you modify your existing single-partition collection to use user-defined performance, your collection will no longer be limited to the 250, 1000, or 2500 RU/s of throughput on your S1, S2, or S3 collection (respectively). Instead, your collection will switch to the new Standard pricing model, and you’ll be charged based on the throughput level you choose, between 400 and 10,000 RU/s. And yes, you can switch back to S1, S2, or S3 collections if you change your mind.

Note that when you change your S1 collection to user-defined performance, you’re getting approximately 60% more throughput for the same price. So if you currently have an S1 collection, with 10GB of storage and 250 RU/s of throughput, which costs you $25 a month, and you change to use a Standard collection with 400 RU/s, it will cost approximately the same, but you’re getting an extra 150 RU/s!

Need more storage or throughput?

Do you need more than 10GB of storage per collection, or more than 10,000 RU/s of throughput? DocumentDB has you covered as well. New partitioned collections, which rely on a partition key to distribute data to among multiple servers or partitions, enable you to ingest larger volumes of data with higher throughput rates than single-partition collections. With partitioned collections, you can now process 10s billions of requests each day with a single DocumentDB collection.

Want to switch your existing collection to use a partitioned collection instead? The DocumentDB migration tools will handle the conversion for you. For instructions, see Migrating from single-partition to partitioned collections.

What are my options for new collections?

Do all these options sound great, but you don’t have a DocumentDB account or collection right now? No worries. These pricing options are available on new collections as well. See How to create a DocumentDB database account using the Azure portal to get started.

If you need any help or have questions or feedback, please reach out to us on the developer forums on stack overflow or schedule a 1:1 chat with the DocumentDB engineering team. Stay up-to-date on the latest DocumentDB news and features by following us on Twitter @DocumentDB.