SharePoint’s interesting hybrid cloud search service

Microsoft announced the availability of SharePoint 2016 Preview yesterday. As expected, This Office server is a tad behind its counterparts in the race to the Office 16 Wave launch expected soon (October 2015 is the bookmaker's favorite) and is on course for a Q2 2016 release. You can’t update this preview version to RTM, so it’s very much a kick-the-tires to see what it does kind of release, which is perfectly fine.

From the perspective of a not-quite-a-SharePoint-guy, the most interesting news is the new cloud hybrid service, available as an update to SharePoint Enterprise 2013 (but not until September 7 - but the documentation is available online to kickstart the planning process). The cloud hybrid service grabs metadata from on-premises SharePoint 2013 servers (and SharePoint 2016, when it is released) and exposes it to the Office Graph database running inside Office 365. The actual documents and lists stay on the on-premises servers, which can run SharePoint 2007, SharePoint 2010, and SharePoint 2013.

Once in the Office Graph, that on-premises metadata is exposed by any application that cares to query the database, including Delve. This is important because a) most large enterprise will run hybrid Office 365 tenants, b) migrating a whole pile of SharePoint data from on-premises servers to the cloud is both expensive and complex, and c) up to now on-premises data has remained invisible to the Office Graph, which removes a great deal of value from an application like Delve because its mission of bringing the most relevant information to the attention of users cannot be met when only half of a company’s infrastructure is visible.

Even if they don’t have access to Delve, on-premises users will also be able to use the unified index to perform searches across both on-premises and cloud resources using standard SharePoint search facilities. Naturally, the on-premises Active Directory has to be synchronized with Office 365 using AADConnect to make sure that user and group identifies are recognized on both platforms.

Another interesting use scenario is when the time comes to migrate workload from on-premises SharePoint to SharePoint Online. Making sure that a common index is maintained no matter where information is stored helps to keep information online as the migration progresses. Of course, this doesn’t mean that a particular document or library might not be available at a particular point in time because data is being moved or the index hasn’t been updated, but the unified nature of the index should mitigate at least part of the complexity involved in the migration.

The Search Foundation is not known for its lightweight hardware requirements and SharePoint administrators will be well aware of the substantial number of servers that are required to ensure reasonable search performance. Part of the benefits promised is that you can eliminate the need for search servers by moving the search index to Office 365. Sounds good, but we’ll have to wait to see the kind of savings that result in real-life deployments.

Microsoft flagged the availability of the hybrid cloud service at Ignite (the presentation and audio is available online - it's a good one!). Some changes are likely to exist between what was described in May and the software now being released, so testing is mandatory to get to know how to maximize the benefits of the new search service.

Moving heavy-duty workload like search operations to the cloud takes advantage of the resources available within Office 365. It also simplifies on-premises operations, so this kind of development makes a lot of sense for those heading towards hybrid deployments.

Those who want more information on this topic can ask questions to Bill Baer, who authored yesterday’s announcement, when he speaks at IT/DEV Connections 2015 on the “The Evolution of SharePoint and SharePoint 2016” at 9:15AM on September 15. Regretfully, no prizes are available for the most difficult question!

Follow Tony @12Knocksinna