Mark Russinovich once again demonstrated Windows Azure virtual machines (VMs)—infrastructure as a service (IaaS). But this time, he was joined by SharePoint MVP Jeremy Thake, who demonstrated unified management of on-premises and cloud (Windows Azure IaaS) SharePoint farms, including replication of content from the on-premises farm to the Azure-based “extranet” farm.
Folks, this really showed off in a SharePoint context what Microsoft discussed at the launch of Windows Azure IaaS three weeks ago. And it’s the “missing piece” in Microsoft’s service offering that will enable heretofore missing support for some critical scenarios. This week, I’ll tell you what I think it means and point you to the recording of the keynote and the short segment where you can see the magic in action.
Many, many organizations are thinking about the cloud. Some, including some of the world’s largest enterprises, are wholly dedicated to moving—quickly—to the cloud. Of course “the cloud” is a problematic term and construct because “the cloud” takes on very different characteristics when it is on premises (a private cloud), outsourced to a provider but in a dedicated form (alternately called a private cloud or a hosted cloud or a dedicated cloud), or hosted on a public service such as (the public cloud). Even within these three types of clouds, there are overlaps and few black and whites.
One common characteristic of clouds other than a purely private cloud is that an enterprise can rarely fully drive architecture and functionality of the service. The cloud service provider sets certain parameters. Parameters may be high-level with a dedicated cloud, or very strict with a public cloud like Office 365 (Shared). Restrictions often include the inability to host full-trust code, to name just one.
When I started using “the cloud” for SharePoint, I procured physical servers from GoDaddy and installed SharePoint on them myself, and fully managed them myself. That wasn’t much of a cloud—it wasn’t as scalable, responsive or elastic as we tend to associate with cloud services today. But from a technical perspective, I had something that’s hard to find today: full control of my service.
There are some providers that deliver both control and elasticity/scalability, but until recently Microsoft has not. Even Office 365 “Dedicated” has pretty significant limitations as to what can and cannot be tweaked and touched by the customer. That left a gap of unfilled scenarios in which an enterprise wants to do something that cloud providers can’t deliver in their “hosted” (or “hosted dedicated”) SharePoint environments, so they are stuck with on-premise architectures, which are often more costly from an infrastructure scenario than desired.
Let’s think about a couple of scenarios: a sophisticated extranet, perhaps consisting of full-trust code for applications; and a desire for an affordable yet responsive disaster recovery environment. Those are two of the scenarios mentioned in the TechEd keynote. I’ve met with a number of organizations, large and small, with needs for DR environments that were too expensive to justify.
Enter Windows Azure VMs—IaaS. Azure enables you to have an infrastructure in the cloud—servers, storage, and networks that span through virtual networks into your on-prem network. On top of that infrastructure you can do darn well what you please!
Mark Russinovich demonstrated the server infrastructure for a SharePoint farm, which consisted of several SharePoint and database servers. Jeremy Thake, Chief Architect at AvePoint, then showed off an example of replicating content from an on-premise intranet to the Azure extranet site. Within seconds, the content was there. Yes, folks, I’m talking SharePoint replication, from on-prem to Azure—live and onstage at a Microsoft keynote. You can see it for yourself at the 1:23 mark of the keynote recording.
Jeremy also briefed me about AvePoint’s support for disaster recovery farms and for storing on-prem backups in cloud storage. That’s where the “light bulb” went off for me, and I plan to use this approach as a disaster recovery environment for our SharePoint farms at NBC in London next month.
Jeremy finished by showing off unified management of on-prem, Azure, and Office 365 farms. With that capability, it means an enterprise now has a full menu of options for infrastructure: on prem servers, private cloud, dedicated cloud, public cloud, and tools to support managing content, compliance, and service delivery. I am sure that some day, Office 365 will offer near-parity with on-prem installations. Until then, when you need to control the service from the Windows layer on up, a solution has now been envisioned and presented.
But let’s also step back from the business-level awesomeness of this short SharePoint “injection” into the TechEd keynote, and away from my awe that it is now possible, technically, to get closer to a “100 percent” solution from Microsoft services. Three other things impressed me about the SharePoint moment at TechEd Europe.
First, it’s extremely rare for Microsoft to allow an ISV onstage during a premier event keynote. They obviously have a lot of faith in the value that this combination of SharePoint, Azure, and the partner ecosystem offers customers. Second, I’m thrilled that the Azure team used a SharePoint example of a multi-tier application. SharePoint is obviously one of the more complex platforms Microsoft offers, with lots of dependencies on other Microsoft technologies. It was brave and laudable for them to use SharePoint in the demo.
Third, I hope this keynote hints at a tighter partnership between Azure and Office 365 in the future. I know Microsoft teams are often internally competitive, but these two services both offer distinct and high-value solutions to customers. If the teams can walk in lockstep, and bring in ISVs to fill in the missing pieces, we will all win.