Eventual Consistency Level
Eventual consistency does not provide any ordering guarantee on reads, but it provides highest throughput.
The replicas eventually converge
Strong Consistency Level
Strong consistency guarantees to always return the most recent committed version of an item.
Bounded staleness Consistency Level
The reads might lag behind by at most K versions or by T time interval. You can configure these parameters.
Bounded staleness offers total global order except within the “staleness window.”
Session
Within a single client session reads are guaranteed to honor the consistent-prefix (assuming a single “writer” session), monotonic reads, monotonic writes, read-your-writes, and write-follows-reads guarantees. Clients outside of the session performing writes will see eventual consistency.
Consistent Prefix
Consistent prefix consistency guarantees that reads never see out-of-order writes.
RTO (Recovery Time Objective) Single Master Any Consistency Level
< 15 minutes
RTO (Recovery Time Objective) Multi Master (Except eventual)
0
Azure Tables partition key
In Azure Tables partition keys separate data into different servers. The same server can store multiple partitions.
Optimize reports with CosmosDB
Use indexes to optimize query performance for on-demand reporting.
Do not use Cosmos DB change feed which is used to trigger Azure Functions based on events.