I was honored to be able to represent AvePoint in the TechEd 2012 Europe keynote on Tuesday morning presenting alongside the legendary Mark Russinovich. I was on stage to demonstrate how our DocAve platform can be used to help improve the benefits of Windows Azure and SharePoint in the cloud.
Our DocAve platform was architected from the beginning to run as a web application and communicate with SharePoint servers through agents deployed to the servers in the farm. This allowed us to support multiple farms from one management platform.
With the new release of Windows Azure Virtual Machines, Microsoft have enhanced the Infrastructure as a Service (IaaS) platform to allow for stateful VMs that now support SharePoint 2010 (FAST is not just yet). Unlike Office 365 SharePoint 2010 Online, you have full access to the servers just like having an on-premise farm, which allows you to install our DocAve Agents and truly customize the farm through full-trust farm solutions.
Our DocAve platform also supports Office 365 SharePoint 2010 Online, but instead of communicating with agents, it talks directly to the client object model and web services end points. This allows you to manage these cloud-based farms alongside on-premise farms.
IaaS for hosting your SharePoint farms in the cloud are not new, the likes of Amazon, fpWeb and Rackspace have offered this service for a while now. For me, the advancements in the Windows Azure Portal and Virtual Machine technology strike very strong competition.
I was very lucky to work alongside Michael Washam, a Microsoft Developer Platform Evangelist team member based out of Redmond. His passion for this technology and the work he showed me made me realize how powerful these advancements are. Michael has a great blog covering all things Windows Azure and I encourage you to check it out. Steve Fox and Paul Stubbs also have some awesome material around this in their book.
Benefits of Windows Azure Virtual Machines
For me, here are the strong points of using Windows Azure Virtual Machines:
- The virtual machines have a variety of core/RAM options and you are charged per day per instance from the time they are created until they are deleted.
- Flexibility to create virtual machines based on templates or VHDs uploaded to your cloud storage subscription through the user interface or PowerShell.
- The benefit of supporting VHDs is that your on-premise Hyper-V VMs can be uploaded to cloud storage and used in Windows Azure Virtual Machine subscriptions.
- The ability to load balance the web front end role servers in your farms into availability sets for true scalability.
- Immediately have public urls for your servers and public IP addresses and the ability to control your end points such as RDP, web applications and others.
- The ability to move your VHDs into a publically available cloud storage location, for others to copy your VHDs to their own subscription and essentially clone your environment in theirs.
Windows Azure, SharePoint and DocAve together
Here are some scenarios we feel Windows Azure, SharePoint and our DocAve platform can help.
Windows Azure could be used to spin up SharePoint servers for a new farm in less than an hour. Business critical site collections could be granularly restored with DocAve from backups in cloud storage. Once the on-premise farms are operational again, the Windows Azure VMs can then be de-commissioned. The main advantage of this approach over traditional disaster recovery farms on-premise is that you are not paying for infrastructure that you may never use.
Cross Farm Management
Administrators can use DocAve Administrator to configure settings or report across multiple Farms at once. This means from an administrators’ aspect, there is no limitations of hosting the farms in Windows Azure VMs rather than on-premise, but with all the cost savings of looking after on-premise infrastructure. Our DocAve Report Center has useful reports such as comparing configuration settings between multiple farms at once to ensure compliance to defined policies.
Development and UAT environments
Developers can spin up development environments in the cloud, without the need to have large development workstations or laptops. They can use our DocAve Deployment Management to move artifacts and content from Production/UAT environments back to their environment to further customize the sites. The benefit here is that you don’t have to restore a whole content database into your development environment, but instead can granularly select down to a list item level. Then once they are complete, can promote the artifacts back to UAT and then through to Production.
In the coming months, I’ll also be investigating using PowerShell with Team Foundation Server to spin up SharePoint Farms to run automated tests as part of the build process to further improve Application Lifecycle Management.
Our Governance Automation product can provide a service catalog to allow business users to make service requests with an approval process for policy based Site Collection Provisioning. It is key that business users don’t have to worry about where their content is and make this transparent to them when they are requesting new project or customer sites and let the defined policies dictate what farm they are provisioned in.
We can also granularly migrate content from one SharePoint farm to another down to a list item level. This means that you can move content workloads from on-premise to the cloud as required. This is a very common scenario when you may start a project workload in the cloud, but maybe after the project has completed, migrate the content down to an on-premise archiving farm for the rest of the lifecycle before it can be disposed of after the required duration.
Hybrid Content Replication
Another common scenario is for organizations to want to collaborate out of their own employees with external users. Traditionally you would poke a hole through your firewall and giving external users access directly to your SharePoint workloads on-premise. The ability to spin up a SharePoint farm that is accessible publically requires less design complexities of the information architecture (site collection hierarchy and security models) to ensure external users can’t see certain internal content.
With our replication product, we can selective replicate content to and from the cloud. In the keynote, I demonstrated how a large on-premise Intranet environment could have authors working on policy documents and videos, which only get replicated if they have been published and approved. The ability to selective replicate content means that rather than isolating internal content and shared content into two different site collections, internal employees can view all of the content in one location behind the firewall and external users access it via another location. We support two-way replication also with conflict resolution functionality for scenarios where content is being edited in both the intranet and the extranet.