WP_Data Management Best Practices

W H I T E P A P E R

www.persistent.com

4.4.1.10What to look for in BI tools/applications?

The front room is the public face of the DW/BI system; it’s what business users see and work with day-to-day. There is

a broad range of BI applications supported by BI management services in the front room, including ad hoc query and

reporting BI tools, dashboarding and scorecarding tools, andmore powerful analytic or mining/modeling applications.

Selecting a BI tool may be a complex process: there are many vendors offering products, none of them solving

completely all requirements (there are very few that combine ad hoc query, enterprise reporting and

dashboarding/scorecarding, for instance), all of them with different strengths and weaknesses (not all of them work

well on OLAP cubes or provide amulti-engine BI semantic layer abstracting a presentation server that can, in fact, be a

logical, federated service), and anyway there is a range of different access needs which no tool meets completely. A

case in point is the recent emergence of new self-service BI systems, see section

below: analysts need data to

solvemore andmore business questions which the existing reports and dashboards can’t solve.

This means that the nuances of selecting a BI tool should pay special attention to the requirements: be as

comprehensive as possible of all the types of calculations; include all critical functionality; obtain all the education you

can before identifying the product candidates, and spend time testing the top candidate systems, involving business

users to understand their perspective, and get their support for the final choice.

4.4.2 Cloud Deployments

Most of the technical architecturematerial for cloud deployments has already been introduced in section

For PSL engagements in this area,

the piece of advice we can give is to first make sure to understand what your

customer’s goals are

. Most of the time they would have made the decision about what type of cloud they need, and

even which vendor. However, requirements may change over time; sometimes there are conflicting requirements, and

sometimes they want PSL to help with the decision, so it is always good to know the list of cloud data warehouse

“advertised benefits” for each type of cloud (review section

to this end), and contrast it to the customer’s needs,

to be in a good position to advise the customer.

Also make, sure you understand the challenges around moving and integrating data in a cloud data warehouse, and

on connecting cloud and on-premises applications, which were also developed in sections

and

4.4.3 Big Data

4.4.3.1 Reference Architectures

Most Big Data projects use variations of some existing Big Data reference architecture provided by data platforms.

Understanding the high-level view of this reference architecture provides a good background for mapping the data

flows with components required and how it complements existing analytics, DW systems. This architecture is not a

fixed, one-size-fits-all approach. Each component of the architecture has at least several alternatives with its own

advantages and disadvantages for a particular workload and type of the data.

It is recommended that architects

start with a subset of the patterns in this architecture, and as they realize value for gaining insight to key

business outcomes they expand the breadth of the use by adding inmore components or tools.