Why to have same service tier for SQL Azure geo-replication?
Currently, Geo-replication can be configured only databases between same service tier... since it is possible that secondary will be used ONLY for DR purpose, maintaining secondary with low config gives the benefit of cost saving. Need your thoughts on this implementation difficulties and possibilities on bringing it to public.
This would be particularly useful with some way to trigger a scale-up when the secondary becomes primary due to a failover. For some applications, we'd accept slow performance immediately after a failover if we know that slowdown is very temporary, but don't want to have to do the scale-up manually.
Peter Jones commented
I think it would be good to be able to replicate to a lower tier, we have an application that requires a high read performance but has very low write performance, generally 95/5 read/write, with 500DTU read perf requirements. But only 10 DTU writes.
Currently to add replication we'd be paying 30x the amount that our performance requirements drive. Even if we had to have a minimum of 50DTU we'd still be saving a huge amount but our data would be so much more secure.
Alternatively we'd be willing to replicate to PRS from P as that would be great for the secondary location which doesn't need the same high resiliancy as it is already a secondary.