WOW! I think we just witnessed one of the biggest VMware-announcements we will remember for a long time this year at VMworld 2016. VMware is going to run a full vSphere SDDC-stack (vSphere / VSAN / NSX) bare-metal on AWS. Let me emphasize on this: It’s running BARE-METAL, so NO nested hypervisor. AWS will be running your workloads on the same vSphere-bits with the same feature sets, you learned to love on purpose-made, dedicated hardware available from all of their AZ’s (Availability Zones).
It gives you a easy way to extend your private on-premises DC to the AWS cloud and have uniform central management plane, leveraging vCenter Enhanced Linked Mode across all your (public & private) vCenter-servers. One of the coolest things which we witnessed was now really moving a VM from your on-premises and totally moving it to AWS with XvMotion-technologies
Deploying your VMC is as easy as following these steps:
- Logging in to VMware-Cloud portal (TBD determined: vmc.vmware.com)
- Selecting your (initial) virtual datacenter size
- Enter your Payment details (CC / VMware-account)
- You’re good to go!
VMware Cloud on AWS gives us a new capability which gives us a solution for a couple of problems we are facing in a vSphere environment. It adds elasticity to DRS, which allows DRS to auto-scale the amount of hosts you are running for your vSphere-cloud running (with set minimum and maximum amount of hosts it’s allowed to spin-up in your VMC). So if DRS detects the cluster being imbalanced it automatically spins up a new ESXi-host, add it to your cluster, move workloads to this new host and get the cluster balanced again within no time. This works scaling the cluster up, but also for scaling the cluster down. So no more waiting 6 weeks for getting a new server ordered and racked up and then finally adding it to the cluster. By then your cluster already overloaded or the load is already gone. Just get the resources you need at any time. Obviously the VSAN-datastore is also extended with every host that’s added to the VMC-cluster. So if you need more capacity on the storage component you just add another host to the cluster.
This reduces the capacity planning you have to do for your vSphere-environment hugely.
Besides capacity management, HA will be leveraged for auto remediation. When a failure occurs on a host within a cluster it will automatically be evacuated (vSphere maintenance mode and VSAN entering maintenance mode and evacuating from the cluster), a new host started and the workloads restarted on this host. If VSAN FTT (Faults-to-Tolerate) is violated another host is added to fix this automatically.
Improved Uptime and Performance
VMware will take care of patching the ESXi-hosts in your VMC cluster and will guarantee you’re having the amount of resources you’re paying for. Cluster size is not affected during these rolling updates / upgrades. VMware will have you on the latest and greatest (stable) version which is available. So no more patch management for your vSphere-environment.
VMC for AWS will be entering the beta phase in the beginning of next year. Can’t wait to do some beta testing. GA will be somewhere mid-2017. Will it be announced GA at VMworld Las Vegas?!
If you’re looking on more information on VMware Cloud on AWS check the following sources: