Our fiscal year-end is November 30 so it is budget time for us. It was five short years ago that we were considering SharePoint to be a “free” solution. We brought our first SharePoint server up on a cheap server, running against the on-board SQL Server instance and we were probably missing a few CALs here and there. Free didn’t stay free for long, within a couple of years, we added a full copy of SQL Server, a new physical server, two copies of SharePoint Portal Server and all the CALs we really should have. We also added Quest’s backup and administration suites, a few Bamboo web parts, a few other web parts and a copy of MetaVis Architect Suit. These products ranged in price from a few hundred dollars for some narrow purpose web parts to several thousand dollars for some of the application suites. Still, I wouldn’t run our operation without them.
I am not going to debate the value equation of SharePoint today. Suffice it to say, that I firmly believe that SharePoint the most cost effective way for us to achieve our content management and collaboration goals. My point today is that even if your main reason for using SharePoint is to save money, you need to seriously consider what a proper SharePoint installation should cost, and you should start with infrastructure.
I’ve already blogged about why you need a test server. Beyond that, you should start to consider redundancy. Even after we started spending serious money on SharePoint, we were running lean; we had SharePoint and SQL Server running on the same box and using on-board storage. As our content grew, we moved storage to an external array. Now that we are upgrading to SharePoint 2010, we are separating SharePoint from SQL Server (virtualizing the front-end and leaving SQL Server on the physical server). We also have a second copy of SQL Server running for our production databases. Having two SQL Servers gives me a rudimentary, albeit manually activated failsafe environment. That doesn’t sound like much of a goal, but achieving it will begin to make SharePoint as reliable as our other network services.
Our failsafe environment will eventually include each of our two SharePoint farms (internal and Internet facing) running on separate virtual servers. The idea is that in the event one virtual server fails, the other could run both SharePoint images. Two copies of SQL Server running on separate servers give us the same level of support for the backend database. We are adding a second storage array this year which will let us store the image files for our virtual servers, and provide a failover in the unlikely event the highly redundant storage array fails.
These are all capabilities I had when we were using file shares for document storage. It was cheaper to support redundancy in that environment, but it is no less important today. We are simply planning and budgeting in a manner that reflects the fact that SharePoint is now a first class object in our network.
#sharepoint #MetaVis #quets #Bamboo #redundancy #planning