Over the last few years, we’ve seen the need for operations management and vSphere to be closely coupled as customers were realizing the value of the two operating hand-in-hand. VMware responded to this customer need and announced today a new vSphere offering – vSphere with Operations Management, or vSOM. The bundle couples vSphere together with the vCenter Operations Management Suite (vC Ops for short). It’s a great addition that fills a gap in our portfolio, but I’m getting ahead of myself, so let’s take a step back to understand the value of this new offering.
If you look at the history of vSphere, it originally started with only the ESX Server hypervisor. The early versions of ESX Server had a simple management web interface (called the MUI) for registering, powering on/off, etc., virtual machines on that ESX Server host. Eventually, as customers started running on the order of dozens of ESX Server hosts, it simply wasn’t feasible to manage each host individually. Customers wanted a central place where they could manage all these hosts and virtual machines (VMs) running on those hosts. Thus vCenter Server was born. Its original charter was basic provisioning tasks such as deploying and registering VMs, powering them on/off, managing snapshots, etc., across a large number of ESX hosts. (As a geeky side note, the primary executable filename of vCenter is vpxd.exe, where the “p” stands for “provisioning.”)
Of course, while vCenter Server’s primary focus was provisioning, it needed to do more than just that. It also needed some on-going operations functions, such as showing the state of VMs, listing events, graphing performance metrics, generating alerts, and more. These functions in vCenter Server were focused on basic data visibility, not on more advanced analysis capabilities. Basic visibility is fine if customers are just starting to dip their toes into virtualization, but as their environments grow and business-critical workloads are virtualized, more advanced operations management tools were needed.
This is exactly why VMware introduced vCenter Operations (vC Ops): the existing operations management tools for virtualization and cloud environments simply weren’t advanced enough. While customers increase the size and maturity of their environments, their operational tools must advance along with them. vC Ops combines multiple disciplines, such as performance, capacity, configuration, compliance, and more. It also works across heterogeneous environments (virtual, physical, and public and private cloud), and provides the advanced features needed by large enterprises while at the same time being easy-to-use for smaller businesses that just want things to work. The best part of vC Ops is that while you can start out using it solely for your virtual environment, due to its support of heterogeneous environments, you can roll it out to your entire datacenter.
We understand that customers are all at different phases in their cloud journey – from basic virtualization to full self-service cloud computing. One thing we know is that operations management is a key capability at all points in that journey. As a result, we’re offering multiple ways for customers to get their hands on vC Ops, and vSOM fills an important gap between standalone vSphere and the complete vCloud Suite:
|vSphere (Standalone)||vC Operations Management Foundation|
|vSOM [new!]||vC Operations Standard|
|vCloud Suite||vCloud Standard → vC Operations ManagementFoundationvCloud Advanced → vC Operations AdvancedvCloud Enterprise → vC Operations Enterprise|
(See the detailed differences between vC Ops editions.)
I want to call out a few important points in the above table. First, as the top row indicates, vSphere customers with current support contracts are entitled to vC Ops Foundation. Yes, you read that correctly – if you are a current vSphere customer, you are entitled to a free edition of vC Ops!
vC Ops Foundation provides core performance monitoring and alerting capabilities for vSphere workloads. It includes advanced analytics that determine typical behavior and show you when metrics are currently outside their normal range. It allows you to define groups based on business requirements (e.g. line of business, production vs. test/dev, or all Windows 7 SP1 VMs) and then to set different policies for each of those groups. That allows you to tailor the alerts for each group based on those policies. For instance, for production workloads you’ll likely want to set alert thresholds more aggressively to be warned at the earliest sign of a problem, while for dev/test workloads you’ll want to set them more loosely or possibly disable them altogether. And again, if you’re a current vSphere customer, you’re entitled to this already. It’s conveniently packaged as a vApp so you can get it up and running quickly. Download it and give it a spin! (Note: vC Ops version 5.6 is compatible with vCenter Server versions 4.1 and above, so please do try out the latest and greatest vC Ops version.)
As the table above also shows, the new vSOM offering includes vC Ops Standard, which builds on the Foundation edition with more advanced performance features in addition to capacity analysis. The additional performance features include the ability to alert you if the behavior of an object has changed. This is a powerful capability that can highlight hard-to-spot problems in the environment, such as when an object is using less resources than normal (which may indicate there’s a bottleneck elsewhere in the system). Consider a set of web servers behind a load balancer. If the load balancer is misconfigured, traffic may only be distributed to a subset of the web servers. vC Ops would notice the unused web servers were getting much less network traffic than normal and generate an alert. This is crucial because the other web servers may not be seeing a lot more traffic than normal, nor would they necessarily be at capacity. So basically, the only early sign of a problem is that the unused web servers are getting very little network traffic. Thus vC Ops can help you detect problems that would otherwise go unnoticed until they created a user-impacting problem.
The capacity features in vC Ops Standard are numerous and include capacity forecasting (when you’ll run out of space), right-sizing (so that VMs are assigned an appropriate amount of resources), and what-if scenarios (to help with planning capacity needs). A common scenario we’ve found with most of our customers is that they’re over-provisioning their VMs (making them too big) and also under-utilizing their infrastructure (too few VMs per host). With vC Ops Standard you can address these common use cases. As with vC Ops Foundation, vC Ops Standard is packaged as a vApp for ease of deployment. You can get it here and I encourage you to check it out!
For customers implementing a private cloud, the vCloud Suite already includes vC Ops. The higher vCloud Suite editions include higher vC Ops editions, with broader and deeper sets of functionality. vCloud Suite Advanced includes vC Ops Advanced, which adds the additional disciplines of configuration (to understand what has changed — and may be causing problems — and rollback capabilities), compliance (such as vSphere hardening guidelines or company-specific rules), cost (cost metering and chargeback), and application dependency mapping (which VMs are talking to which, what applications are running in which VMs). It also offers support for heterogeneous environments and more advanced capabilities like custom dashboards and the ability to plug-in “adapters” that connect to and pull data from vFabric Hyperic, EMC VNX, MS SCOM, external SQL databases, and many other data sources. vCloud Suite Enterprise includes vC Ops Enterprise, which builds on everything just mentioned with performance, configuration, change management and compliance for guest OSes and applications. These guest-level capabilities are offered both for vSphere environments as well as other platforms. Both vC Ops Advanced and Enterprise are available here.
In closing, we’ve developed strong provisioning capabilities in vCenter Server and now we’re complementing that with strong operations capabilities in vC Ops. In my opinion, they’re two sides of the same coin: vCenter Server is used to build your environment and vC Ops is used to run it. Because they are so complementary, we think every vSphere customer should use vC Ops and indeed, as I mentioned, all vSphere customers are entitled to the vC Ops Foundation. And with vSOM, you get even stronger operations management functionality from Vmware.
We’re excited to be able to help our customers better manage their vSphere environments and want to hear your feedback. Have you had success in pairing operations management with vSphere or is this something you plan to do in 2013?