In 2016 VMware made big news with our Predictive DRS (pDRS) announcement and analysts and customers have been buzzing about it ever since. Predictive DRS is a game changer and helps you in the fight against resource contention. As you know, when contention occurs, applications slow down and your users are affected. Successfully mitigating resource contention is one of the most critical issues faced by any administrator of a virtualized environment.
The newest release of vRealize Operations 6.5 improves pDRS by allowing it to scale for ANY environment. Previously we recommended running pDRS on clusters with under 4,000 VMs. That “limitation” has been removed, so pDRS can now be used in your largest clusters.
I get really animated when I tell the Predictive DRS story. I love the simplicity of it all. How the two previous resource contention methodologies (Reactive and Balance) have their pros and cons. How they are augmented by the new Predictive method. How together these three methodologies provide a safety net for your workloads, ensuring they get the resources they need.
I think this new video explains it best: VMware Predictive Distributed Resource Scheduler
In a complex datacenter, all three risk mitigation methods are required to fully manage contention. VMware DRS and vRealize Operations offer you all three, in cohesive one solution.
For a more in-depth look at how DRS & pDRS function, I recommend these blog posts.
https://blogs.vmware.com/vsphere/2015/05/drs-keeps-vms-happy.html
https://blogs.vmware.com/vsphere/2016/05/load-balancing-vsphere-clusters-with-drs.html
https://blogs.vmware.com/management/2016/11/predictive-drs.html
https://blogs.vmware.com/management/2017/02/setting-record-straight-drs-pdrs-workload-placement.html
