Do you know that DRS can simplify your administration?
Back in 2006, when DRS was released to simplify the administrator's life and current version totally changed and come with more advanced as well as enhanced features.
Now, the new DRS version was rewritten with more fine coding to manage the workload in a systematic way. this article gives you more insight into its new features and algorithm
.
Old DRS
It was focused on cluster state, checking whether it needs a rebalancing & run every 5 min, and if the DRS logic determine it could improve the cluster balance, it would recommend and execute a vMotion else take no action.
The New DRS
The biggest change in the new DRS is that it no longer balances the host load directly. instead, it improves the balance by focusing on the metrics (virtual machine happiness) and it runs every minutes rather than 5 minutes in old DRS. This result in overall betterment performance of the workload.
This new DRS process logic qualifies the virtual machine happiness by using DRS score. this more means on execution efficiency of the virtual machine and score values are range from 0 to 100%. which again divided in four bucket 0-20%, 20%-40% , 40%-60%, 60%-80% and 80%-100%.
DRS will try to maximize the execution efficiency of each virtual machine in the cluster while ensuring fairness in resource allocation to all virtual machines
To Conclude
DRS is providing a very fundamental and important functionality as part of our virtual infrastructures since its first release. These important enhancements in vSphere 7 mean that workloads will perform as close to optimal as possible, while ensuring that hardware is used very efficiently, and help vSphere Admins add GPUs and hardware accelerators to workloads.