Sharepoint - An Introduction

Editor's Notes

• #3: Record Principle & Rules
• #4: Susan With your evaluation of SkyDrive Pro, this becomes a natural extension of your Portal where you can share, store and collaborate on files. This approach lowers storage costs while still providing you with a seamless way to collaborate.
• #7: -> Cache service: based on appfabric. Not supported from Azure yet. Activated on SP server level. (no technical detail available yet) -> Request management: Optimize WFE usage with optimal redirection of requests. Configured on Web app level with routing rules and health index.
• #8: Here’s an example of what we’ll call a small farm. It is not a dedicated search farm so you can see we have other application servers to run all the other SharePoint service applications. For search specifically we’ve created separate servers for the index and query components, and another set for all the other search components. You’ll notice that we have added a replica for the index partition so we have high availability, and we’re keeping the query role with the index role for maximum query performance. You’ll also notice that all of the search components are highly available because they’ve also been created across host servers. So in this case all of the components are running on virtualized servers, but those virtual images are split across two hosts so if one host goes down, we don’t lose all of any one of the search service components. ********************************** From http://officecpub/Teams/itpro/Documents/O15%20Models/sp-2013-enterprise-search-model-.vsd From Olaf’s email: Note that the 4x VM need, is really a mapping of the need for 16 CPU cores for sufficient processing horsepower. On Win7 hosted VMs, that transfers to 4x VMs. Using physical servers or Win8 hosted VMs, there is no practical max CPU core per machine limitation, thus you can get away with only 2x machines as in O14.   So 2x dedicated _machines_  for search is still supported, you although need to split those out into 4x VMs if running with Win7 has the host OS for the VMs.   Topology diagrams are here: http://office/15/performance/IEG/Shared%20Documents/Productivity%20Search%20topologies%20v2.vsd
• #9: In this case we’re bumping up the size of our corpus to 40 million items. We now have a dedicated search services farm, but in this diagram allow for the fact that if you wanted to have one big farm you could still do that – you would just add additional servers for the other service applications and add the service application databases to the SQL Servers. We’ve also increased the number of components. We now have four index partitions, with each partition having a replica for high availability. We still have two query processing components because that’s an adequate starting point unless we find we have enough load to require more. We’ve also bumped up the number of content processing components we have, to four. We’re only using two crawlers because relatively speaking that’s a lighter weight process retrieving the content, than actually processing it. ************************************** From http://officecpub/Teams/itpro/Documents/O15%20Models/sp-2013-enterprise-search-model-.vsd
• #10: Here’s our large example, which is 100 million items. Now we’re clearly in a dedicated search services farm. You can see that we now have 10 index partitions, and each one has a replica for high availability. We’ve also increased the number of query processing components up to four to handle the increased load one might expect from that many items. On the content processing side we’re still just using two crawlers but we’re up to six content processing components, and we’re also up to six analytics components. Remember analytics activity is based on search activity, as well as usage activity and how many items have been crawled, so as our corpus size increases so will the demand on analytics. Again though, how many crawlers or any OTHER component you are going to need is going to be completely dependent upon how big your corpus is, how frequently it changes, what your freshness requirements are for indexed content, how much query load you have, etc. **************************** From http://officecpub/Teams/itpro/Documents/O15%20Models/sp-2013-enterprise-search-model-.vsd
• #16: From OneNote with Eray
• #19: Previous slide was about the SharePoint hosted apps, what’s the difference to provider hosted?
• #20: Apps for SharePoint can be self-hosted by a developer, auto-provisioned by Azure, hosted by SharePoint, or a combination of these. All can leverage SharePoint components. Custom code can only be run in developer self-hosted or Azure auto-provisioned hosting options. The app is hosted in the cloud. Windows Azure and SQL Server Azure components are provisioned automatically when an app is installed. (Available for SharePoint Online Service only.)