VMware vSphere update is out, but before looking into each of the features and understanding how they’re upgrading, we need to know what they’re targeting.
By early 2016, VMware decided to stop being called an industry leader who has past his prime and decided to prove that its powerhouse product ‘vSphere’ had not reached its maturity stage. In the vSphere product lifecycle, the scope for blockbuster success was closing in. On the other end, Amazon was relentlessly dominating the Cloud industry.
By October of the same year, a news broke out about the partnership between the long-term rivals VMware and AWS. VMware technology could be run on Amazon Cloud and this lead to a new initiative called VMware Cloud-on-AWS. VMware had made a move to maintain its footprint on all new technologies.
This move was to target four kinds of user bases:
- The ones who looked to expand their data centers into public clouds for disaster recovery
- The ones who wanted to expand their geographic capacity footprint
- Those who wanted to reduce their data center overhead by running their workloads in the cloud, and finally
- The smart ones, who wanted to utilize as many services as possible from both VMware and AWS
It took another year, until 2017, for them to deliver on their promise.
But, from then on, they have defibrillated their vSphere and tweaked it to suit their Hybrid Cloud strategy.
VMware vSphere 6.7 is just a step in that direction
It is not a big release or a big bang update. But it does send a clear message that they’re going to create a consistent infrastructure for users across all platforms.
They’re making the management easier, security more effective and performance drastically improved. According to them, the vCenter server (the keystone of Hybrid Cloud Strategy) is 200 percent as good as the previous version ever was, memory usage reduced by 300 percent and Distributed Resource Scheduler faster by 300 percent.
vCenter Server 6.7
vCenter Server, for the uninitiated, is a management and monitoring tool with which you can control your entire vSphere infrastructure. All your VMs and ESXi hosts can be managed from a single console.
This vCenter Server was initially deployed on Windows and Linux servers. But during the release of vSphere 6.5, they had announced that they were pulling the plug for Windows deployment support. It was more of a product metamorphosis than pulling-the-plug.
With vSphere 6.5 Update 1, the VCSA (vCenter Server Appliance), the evolved form of vCenter Server, had become a prominent member of vSphere. With the latest release, they have decided to strap in and accelerate every parameter on all fronts.
For starters, the entire deployment model is simplified and enhanced, thus reducing the number of nodes to maintain. Once upon a time, it was a SSO to manage all the services. But during vSphere 5.1, they decided to spice things up with SSO and distributed services.
Later, they realized things were easier before and turned their car around, this time with the Embedded Platform Services Controller (PSC) – a single sign-on that will empower you with as much as 15 deployments in a single domain. Load balancer is not required for Native vCenter Server High Availability.
Those who are using the server on Windows, need to migrate to VCSA and this is simpler than ever. You migrate all the history and performance data to the appliance in two ways:
- Start the appliance after the configuration data has been migrated and continue other data migration in the background
- Start only after all the data is transferred (even the estimated time is displayed). Upgrades can be only from 6.0 or 6.5 and not from antiques like 5.5
vSphere Client has been upgraded to HTML5-based Clarity UI Web Client from the unstable, frequently-patched and, frankly, a shoddy Flex-based interface that looked straight out of a spreadsheet. Apparently, changing to HTML5 has been in their plans for sometime, but the HTML5 tools had not caught up then. Now, this web client can do more than the Flex-client could do.
There is also an Update Manager cut into this that simplifies host patching and upgrading, by reducing the number of steps involved and allowing for a pre-check option that verifies the cluster readiness before workflow initiation.
A modern server hardware that is rigged with hundreds of gigs of RAM takes up several minutes (that seem like several hours) to initialize, boot and run some basic troubleshooting tests. 6.7 has fastened this process while also offering another feature called Quick Boot. When you load an ESXi, typically the entire hardware is restarted. Restarting the entire hardware in a server is the last thing to be done to speed up a process. Quick Boot restarts the kernel without disturbing the hardware, thus achieving boot within a couple of minutes. But this feature is not compatible to all platforms. Checkout Quick Boot compatibility list for more info.
There is also a plug-in available for vROPs (vRealize Operations) Manager in vSphere Client. vROPs is also a monitoring tool, which provides information about the performance and capacity of the vSphere infrastructure. By adding this Plug-In, you’re welcomed in your vSphere client with a dashboard of vROPs, delivering segment-wise analysis of resource allocation, capacity optimization, risks and health.
Backup & Restore
The File-based backup of vCenter which debuted in 6.5 has taken the stage with a separate tab. Scheduling options and retention policies along with a Restore workflow has enriched the backup policy of VCSA.
While talking about backups, it is necessary to mention the number of new APIs released related to backup scheduling, monitoring, deleting and initiating. They’ve also updated APIs for VCSA management options. An API for Instant Cloning of a VM sharing the same configs, memory and storage, which started out as Project Fargo in early 2014, has been released.
Their boosted support for NVIDIA vGPUs will help designers and engineers along with IT admins to run the GPU-accelerated systems for advanced operations like Machine Learning and AI using the new Suspend/Resume capabilities for VMs.
In storage, typically the cheaper devices have higher latency. DRAM (RAMs used in computers) has faster read/writes but burns a hole in your pocket. 6.7 has released support for Persistent Memory (PMEM). It means, you can add a NVM (non-volatile memory), that costs somewhere in the range of Flash, as a PMEM datastore and use it at a speed somewhere in the lower end range of DRAM, with no guest-OS or app changes whatsoever. This improves app performance because of the direct access to hardware.
There is also improved support for RDMA (Remote Directory Memory Access) over Converged Ethernet (aptly called “Rocky). It essentially means, the memory of one computer can be accessed directly by another memory from another computer without involving any of the two OS, over a converged ethernet network.
This, along with Fibre Channel over Ethernet and iSCSI extension allow for improved performance by bypassing kernel and OS. This reduces the latency.
The final major aspect of the 6.7 upgrade is the TPM 2.0 support. Trusted Platform Module in your computer stores all the encrypted passwords and hash values. This TPM needed to support VMs. VMware, as it always has, decided to solve this by virtualizing – this time the TPM. TPM operates in a VM, pretty much the way it operates in a physical machine. Except this time, to secure the data stored there during VM migration to another data center, VMware chose to encrypt this nvram file with VM Encryption rules.
Apart from this, another update of AppDefense has been released. This is a result of ideation from the team behind NSX. After the release and early signs of success of NSX, the team decided to enter security and come up with a solution for encryption in data centers. They started working on Project Goldilocks by 2016, which turned out to be the core behind AppDefense.
The idea was fairly simple. Why build systems and then secure them when you can enforce all the necessary security measures while building it? An app should therefore, before implementation, be given all the required permissions based on its expected behavior. So, when it is used, it will be secured and cordoned off in such a way that it can access only certain ports, run certain executables, explore certain network infrastructures and so on. And all these behaviors are monitored and reactive measures such as Quarantine of VM, VM Suspension, etc. are available to enforce during a breach.
All the updates offered by VMware vSphere 6.7 have been engineered to support workloads that run anywhere, from on-premise data center to cloud. Their Hybrid Cloud strategy has been strengthened from the foundation by expanding their technologies everywhere and to everyone.Like what you read? Rate us