Keep up with the demand of increasing data volume by partitioning databases on the cloud
Directing a query to a single partition uses a fraction of the database's resources, allowing for speedy query times and lower per-query pricing.
When determining the best approach for data storage in the cloud, you can face several challenges and questions.
First of all, scaling a cloud application from initial development (where there is zero traffic) up to handling thousands of requests per second is a difficult engineering problem to solve. The data layer bears the brunt of this growth, having to handle an ever increasing data volume while servicing incoming queries in a timely manner.
One way of keeping up with demand is to partition a database into smaller pieces, where each partition contains data that belongs together, such as readings from the same IoT device or orders from the same user or products belonging to the same category. Directing a query to a single partition uses a fraction of the database’s resources, allowing for speedy query times and lower per-query pricing.
You can address these challenges by using partitioned database capabilities based on the open-source Apache CouchDB™ project. Cloudant is a JSON document store, built on CouchDB, that is a highly-available distributed system for storage and retrieval of data.
You can try out Cloudant as part of IBM Cloud. Get advice about partitioned databases, part 1 in a Cloudant blog series and then learn even more in the Cloudant documentation.