Early this month, VMware announced vSphere 6.0, the latest version of the most popular and powerful enterprise virtualization platform. As a longtime VMware shop with a number of VCP certified professionals, we’re excited for a number of the new features included in the latest release.
According to VMware’s CEO, vSphere 6.0 has over 650 new features. Green House Data doesn’t yet have a timeline for upgrading from 5.5 just yet (6.0 hasn’t received General Availability), but here’s what we’re most excited about.
Many of our cloud customers use the web client login rather than a desktop client to manage their vSphere powered cloud infrastructure. In 6.0, numerous improvements are coming to the web client, including a 13x faster login, faster and flat menus, general performance increases of about 50%, integration with the VM Remote Console, and more frequent refreshing.
While the client is still Flash-based (much to the chagrin of many who have called for an HTML5 version), there have been some improvements to the UI in general, with a number of ways to maximize screen estate in addition to performance increases. Alarms and Work in Progress panels can live in small expandable boxes on the side, and the dock is now placed on the bottom of the screen.
There has been a lot of buzz about this new feature as it could radically change how storage is consumed and managed in vSphere. In vSphere 5 and earlier, all virtual machines inherited the policy and abilities of the associated datastore. That means the storage method dictated the storage characteristics of VMs.
vSphere 6.0 will introduce virtual volumes, storage APIs, as a way to allow all VMs to have their own storage policies, freeing administrators from having to frequently fight with LUNs (logical unit numbers, used to assign hard drives on a network), file systems, and storage policies. This is an implementation of Software Defined Storage (SDS), but depends on partnerships with storage vendors who will support this technology. However, many major vendors are on board (Dell, NetApp, IBM, EMC, Nimble Storage, Solidfire, and others).
Ultimately, this means administrators simply assign a pre-defined storage policy to each VM that controls the performance and capacity, without needing scripts or diving into individual LUNs and volumes. Results? Faster snapshots, clones that use less storage space, and hardware accelerated management.
Previously, fault tolerance in VMware kept a VM clone in very close sync with a live VM so that if one of the two failed, no network connections would be lost and not even a single CPU cycle would be missed. For critical applications, this was a great feature! But it was limited to a single vCPU, and couldn’t be used with snapshots.
In vSPhere 6, fault tolerance has spread to four vCPUs per VM, with higher memory allowances as well, so it can be used with larger resource intensive applications. No user snapshots are supported, but VADP (vStorage API for Data Protection) snapshots can be used for backup and replication.
Any major release is going to bring with it performance and capacity increments as well. For 6.0, each cluster can now handle 2x as many VMs (8,000 vs 4,000), as well as 2x as many hosts per cluster, from 32 to 64. Each host can handle 50% more CPUs, up to 480 now, and up to 12 TB of memory.
Each VM can also now use up to 128 vCPUs and 4TB of RAM, previously maxxing out in 5.5 at 64 vCPUs and 1 TB of RAM. This should go a long way towards enabling Internet of Things and big data applications, as enormous datasets can be handled by a single virtual machine, with simplified disaster recovery and backup.
Instant Clone is a new feature that enables near-instantaneous cloning of live virtual machines. Combined with improvements to vMotion, administrators can now live clone machines across vCenter servers and across longer distances. Depending on the network connection round-trip times, Green House Data could clone live virtual machines across our data centers, from Oregon to Wyoming to New York, almost instantly! This is a great feature to pre-empt disaster recovery plans, allowing administrators to move VMs around geographically diverse data centers quickly and easily in anticipation of future downtime.
We’re happy to see VMware iterating and improving their virtualization platform, and are looking forward to delivering these advantages to our customers as soon as we can.