From the course: Microsoft Azure Cosmos DB Developer Specialty (DP-420) Cert Prep by Microsoft Press
Unlock this course with a free trial
Join today to access over 24,800 courses taught by industry experts.
Calculate and evaluate throughput distribution based on partition key selection
From the course: Microsoft Azure Cosmos DB Developer Specialty (DP-420) Cert Prep by Microsoft Press
Calculate and evaluate throughput distribution based on partition key selection
Let's revisit the idea of the Request Unit. The request unit or RU is the standardized metric of performance for Cosmos DB, and the way that Microsoft generates what 1 RU is, is how much they'll charge you in your currency for a point read for a one kilobyte item that's 1 to 1. And I also mentioned that the RU value is deterministic. So each time you run an identical query, you can expect an identical RU charge. Now, a point read, like I said, is the very most efficient way to do a query because you know what the ID and partition key is for that object. In other words, the partition key means that you're in the correct container and because you know the ID value of the document in advance, there's no need for indexing or scanning. You can go right to the appropriate partitions and gather up that file. Now you should know that writes cost approximately 5 times as much as reads do, so it's a more expensive operation. Some other…
Contents
- (Locked)Learning objectives37s
- (Locked)Choose a partition strategy based on a specific workload2m 24s
- Plan for transactions when choosing a partition key9m 18s
- (Locked)Calculate and evaluate throughput distribution based on partition key selection12m 41s
- (Locked)Design partitioning for workloads that require multiple partition keys25m 19s
- (Locked)