Scaling SharePoint within a global organization of 100k plus users isnt a simple task. Generally your dealing with several factors such as:

  • Local laws regarding data and privacy
  • WAN constraints
  • Localization
  • Storage requirements
  • Data centre constraints
  • Operational constraints
  • Service level expectations

To make it even more complex throw in a healthy dose of politics and a new CIO/IT director(s) looking to make their mark.

When designing solutions for such complex environments its important to balance user demands such as:

  • Features and functions
  • Full administration and customization access
  • Spin on a dime service

With IT operations that:

  • Wants a life outside of work
  • Are in fear of users with too much power
  • Is on the hook for support

On many occasions my role is to develop a roadmap that provides the business and IT with a plan for establishing a collaboration service based on digestible chunks – balancing the realities of resources and time. For example, define an SLA, establishing a lab to ready IT skills, impact assessment (Data centre, WAN, Operations) and model solutions for select business units as proof of concepts – key steps required before deploying to their global user base.

Though some would argue that planning can lead to analysis paralysis, taking unnecessary risk leads to business impact and people getting fired. Its important to note that I’ve worked with several companies that have failed miserably at their past deployments for a variety of reasons, usually a result of lack of:

  • Experienced staff
  • Funding to staff the project team properly
  • Typical project management horror stories

My analogy is often one of building a house. Do you pay a contractor whom has never built a house to build one for you? If you like taking wild risks then go for it. If not then you should look for a contractor that has built homes for a living.

So what are the top pain-points?

  • WAN – where is bandwidth limited and what is the plan for addressing it. More? Accelerators? Small farms?
  • Backup and restore – how mature is my SQL management capability? What’s my recovery window? Can I use existing infrastructure?
  • Capacity management – how do I control growth? What tools do I use? How can I look it down?
  • File shares and Public folders – what do I do with them? What do I migrate? How do I migrate?
  • SSO – Do I use AD? LDAP? Will I need MIIS?
  • Feature management – what features does the business require? whats the licensing impact? What’s the operational impact?
  • Data centre – do I have space? Power? AC? Staff? Monitoring and management tools? Must I virtualize? What’s the impact?
  • Taxonomy/Information architecture – how to establish and maintain an architecture that wont grow out of control and eventually become useless?
  • Custom applications – how do I address custom applications? Whats the impact to my servers, network and storage? Operational impacts?
  • Help desk – what sort of calls with they handle? How do I route calls and to whom? How much self service can I expect to make use of?
  • Compliance – how do I remain compliant? How do I prevent users from misusing SharePoint for records?

I’m sure more comes to mind but I believe these are the top issues. So how do you address these? Organizations commit to providing a collaboration service much like they did when they deployed a messaging service. Its scary at first but once they get past the fear and denial, they get on with it.

Advertisements