Over the last 5 years I have observed Microsoft persevere
with the notion of Self Service BI over a series of conferences as far back as
SQLBits V in Newport. The release of SQL Server 2012, improvements in Excel and
the integration with SharePoint 2010 is making this a reality.
Business users are now empowered to create their own BI
reports through a number of different technologies such as PowerPivot,
PowerView and Report Builder. This opens up a whole new way of working; improving
staff productivity, promoting efficient decision making and delivering timely
business reports.
There is, however; a serious question to answer.
What happens should any of these applications become
unavailable? More to the point, how would the business react should key
business users be unable to fulfil reporting requests for key management
meetings when they require it? While the
introduction of self-service BI will provide instant access to the creation of
management information reports, it will also cause instant support calls should
the access to the data become unavailable.
These are questions that are often overlooked when a
business evaluates the need for self-service BI. But as I have written in other
blog posts, the thirst for information is unquenchable once the business users
have access to the data. When they are unable to access the information, you will
be the first to know about it and will be expected to have a resolution to the
downtime as soon as possible.
The world of self-service BI is pushing reporting and
analytical databases to the tier 1 application level for some of Coeo’s
customers. A level that is traditionally associated with mission critical OLTP
environments. There is recognition that by making BI readily available to the
business user, provisions also need to be made to ensure that the solution is
highly available so that there is minimal disruption to the business.
This is where High Availability BI infrastructures provide a
solution.
As there is a convergence of technologies to support a self-service
BI culture, there is also a convergence of technologies that need to be
understood in order to provide the high availability architecture required to
support the self-service BI infrastructure. While you may not be the individual
that implements these components, understanding the concepts behind these
components will empower you to have meaningful discussions with the right
people should you put this infrastructure in place.
There are 7 worlds that you will have to understand to successfully
implement a highly available BI infrastructure
1.
Server/Virtualised server hardware/software
2.
DNS
3.
Network Load Balancing
4.
Active Directory
5.
Kerberos
6.
SharePoint
7.
SQL Server
I have found myself over the last 6 months reaching out to
knowledge that I learnt years ago when I studied for the Windows 2000 and 2003
(MCSE) Microsoft Certified System Engineer. (To the point that I am resuming my
studies for the Windows Server 2008 equivalent to be up to date with newer
technologies) This knowledge has proved very useful in the numerous engagements
I have undertaken since being at Coeo, particularly when dealing with High
Availability Infrastructures.
As a result of running my session at SQLBits X and SQL
Saturday in Dublin, the feedback I have received has been that many individuals
desire to understand more of the concepts behind the first 6 “worlds” in the
list above.
Over the coming weeks, a series of blog posts will be put on
this site to help understand the key concepts of each area as it pertains to a
High Availability BI Infrastructure. Each post will not provide exhaustive
coverage of the topic. For example DNS can be a book in its own right when you
consider that there are so many different configuration options with Forward
Lookup, Reverse Lookups, AD Integrated Zones and DNA forwarders to name some
examples. What I want to do is share the pertinent points as it pertains to the
BI infrastructure that you build so that you are equipped with the knowledge to
have the right discussion when planning this infrastructure.
Next, we will focus on the server infrastructure that will
be required to support the High Availability BI Infrastructure, from both a
physical box and virtualised perspective.
Thanks
Chris
