Is there an autoscale option instead manually scalling the vCores?
@ArvindShyamsundar-MSFT2 ай бұрын
There is no service level support currently for auto-scaling an entire elastic pool. However, we can have a custom approach to monitor elastic pool resource usage and accordingly implement scaling in an automated manner, albeit managed outside of the SQL DB service itself. I cannot seem to post links for some reason. If you can please search for a blog article called "Autoscaling with Azure SQL Hyperscale" authored by my colleague Davide, he shows how you to do this for single databases. Equivalent approach can be considered for elastic pool as well. In his article, he uses sys.dm_db_resource_stats to monitor resource utilization for a single DB. You can use the equivalent, sys.dm_elastic_pool_resource_stats to monitor the elastic pool level resource utilization.
@bsoverns2 ай бұрын
Where can I find information on the MBps?
@ArvindShyamsundar-MSFT2 ай бұрын
I'm unable to post links for some reason, but if you can search for "Resource limits for elastic pools using the vCore purchasing model" and specifically within that page, the sections therein for "Hyperscale", you will find the answer you need. In summary, Max log rate per pool is 125 MiB / sec. However, each individual database within the pool cannot exceed 100 MiB / sec.