Is there any document clearly telling MAX RPO and Availability attained through SQL Server AlwaysOn Availability Groups/ Mirroring? i.e. 1 min and 99.99% like that?
I don't believe that generic answer exists. As usually with sql server a lot depends on application. Say you have long transaction and when it is almost over, something goes wrong and you fail-over. Until new primary node rolls back the transaction database is not available for query and you have downtime. Also consider cluster (if we talk about availability groups) that has its own time to recover. The only way (IMHO) is to carefully monitor your application and environment and do testing with real scenarios. Even the simplest scenario when you have to patch a node and need to do a planned fail-over gives you some down time. as for RPO - sync AG has zero, async very much depends on hardware and load (the more replica behind primary node the larger loss you may have). RTO also depends on the application.