Edit

Improve performance and optimize costs when upgrading to Azure Cosmos DB API for MongoDB 4.0+

  • Article

APPLIES TO: MongoDB

Azure Cosmos DB API for MongoDB introduced a new data compression algorithm in versions 4.0+ that saves up to 90% on RU and storage costs. Upgrading your database account to versions 4.0+ and following this guide will help you realize the maximum performance and cost improvements.

How it works

The API for MongoDB charges users based on how many request units (RUs) are consumed for each operation. With the new compression format, a reduction in storage size and query size directly results in a reduction in RU usage, saving you money. Performance and costs are coupled in Cosmos DB.

When upgrading from an API for MongoDB database account versions 3.6 or 3.2 to version 4.0 or greater, all new documents (data) written to that account will be stored in the improved compression format. Older documents, written before the account was upgraded, remain fully backwards compatible, but will remain stored in the older compression format.

Upgrading older documents

When upgrading your database account to versions 4.0+, it's good idea to consider upgrading your older documents as well. Doing so will provide you with efficiency improvements on your older data as well as new data that gets written to the account after the upgrade. The following steps upgrade your older documents to the new compression format:

  1. Upgrade your database account to 4.0 or higher. Any new data that's written to any collection in the account will be written in the new format. All formats are backwards compatible.
  2. Update at least one field in each old document (from before the upgrade) to a new value or change the document in a different way- such as adding a new field. Don't rewrite the exact same document since the Cosmos DB optimizer will ignore it.
  3. Repeat step two for each document. When a document is updated, it will be written in the new format.

Next steps

Learn more about upgrading and the API for MongoDB versions: