We have four virtual machines, each with SQL Server for develop, testing and QA. All of the Virtual machines are running in the same storage account. A developer restarts one of this VMs because of performance. After restart, for some reason, one of the databases running in the VM start in Recovery. This was not the problem. The problem was the recovery mode last 4 hours (30 GBs MDF + LDF) - Checking Resource monitor I saw IO write bytes between 40kb and 500kb per second.
To test we start to copy files between folders and the copy throughput (measured by Windows Server Copy Dialog) was between 1mb and 2.5 mb per second.
Is this a normal measure for this service? DO we need to move our virtual machines to a datacenter with premium storage support?
There's no alert for South Central US since August!!
Regards
And please forgive my english writing :D
Javier Schamber