Azure Cosmos Db
- Refer Here for Request units
- Refer Here for conversion from vcore to request units.
- Autoscaling vs manual provisioning Refer Here
- Create a cosmos db and navigate to the data explorer
- mongo db
Traditional vs Serverless
- In traditional approaches we create/use server/vm/container to run our workload and we add code/data to it. The server needs to be up all the time. In cloud models we pay hourly charges
- Serverless is a technology in which we just focus on code/data and which when a request is made the server is automatically provisioned by cloud service providers which runs your code/data and you will be charged per seconds/milliseconds based on time taken to process requests.
- The code which we run in these serverless technologies is called as functions
(AWS => Lambda, Azure => Azure functions, Gcp => Google functions )
- Then this technology was adopted for data related applications as well
- These required to persist data.
- Problems with Serverless
- No Standards yet
- Not portable
Next Databases
- Cache Databases
- Azure Synapse
- Azure Data Factory
- Azure Databricks