Once Upon a Time we Thought Analytically… (Part 2 of a 3)

By Guest Blogger Kemal A. Delic

Looking back into the recent history of IT industry developments, one can easily observe the domination of hosting platforms—representing an intricate infrastructural entity combining technological systems and a clever business model. The most important feature of the dominant, global platforms is the ability to scale smoothly. You provide the platform and they will come (the old chestnut). As we have established our Radix platform we faced the problems of performances and more accurate capacity planning, as the number of customers has been reaching many hundreds very rapidly. Staffing with human operators will be expensive and not so rapid—as monitoring and management analytics was the way to go.

FROM MANAGEMENT ANALYTICS TO ACTIVE ANALYTICS

So, it was clear to us that we needed to provide first exhaustive monitoring subsystem to feed management analytics supporting operating personnel with timely and accurate data synthesis and interpretation. An important insight about different decision-making styles as intuitive for the corporate executives, rational for managers and analysts and routine for employees was combined with volume and type of data captured in Fig. 1. – from our Intelligent Enterprise journal article.

IE_Fig_1

Fig. 1 Decision style coupled with type and volume of relevant data

Another important insight is captured in the same article in  Fig. 2. suggesting stratification of the enterprise architecture into three distinct layers to deal clearly with defeating enterprise IT complexity.

IE_Fig_2

Fig. 2 Layered Enterprise Architecture able to cope with complexity

So, the first layer consists of enterprise events, which require real-time capabilities, very low latencies and the ability to cope with event avalanches. A transactional layer has near real-time capabilities where no lost transaction should happen and the system must be able to recover fully and accurately from any outages. Third layer analytics is fed from the previous two layers and are typically non-real-time, as it includes long-running simulations and uncertain querying of knowledge bases. Each layer contains appropriate storage types and delivery portals designed for specific user needs. The next improvement was developing Active Enterprise Article (OVUA article) in our joint work with famed HP Labs.

active_analytics_

Figure 3. The most intensive event layer—fully automated

Previously developed layered architecture is further developed in the lowest operation layer, providing self-managing IT architecture which we have described in our IEEE article in more detail. Important to notice and make the distinction among business analytics, operational analytics and system analytics. They have different role, nature and characteristics—which is an important architectural distinction.  This also provides insight into how software is designed, a specific view for a specific persona, not a hodgepodge of functionality on a single UI element; taking role-based access control to a new level.

As always, we have ensured intellectual property as dozens of patents, copyrighted software and published articles and papers. From this patent portfolio, we select this European patent on security analytics as street legend says that they are much more difficult to obtain and have higher value and quality.

ONCE UPON A TIME..,

We thought (at that time) that our analytics architecture would be the best and most natural extension of the famed HP Open View product suite. The commercial reality and changed world around us have made this belief the pipedream. Acquisition of the external company had precedence to any (lengthy) internal developments which have seen our plans wrapped-up and canceled.

One more time, we have learned that having good ideas is just the humble beginning of the long and winding road of getting the budget, deserving the trust and long-term commitment of senior company leadership. Ideas we have captured and shared are having even today certain technical sense and business value—even after more than ten years passed from our dreaming.

The next and last posting will be our postulate about the rising science of service systems (2006) which is better known today as the Digital Economy. Stay tuned…

REFERENCES

[1] Intelligent Enterprise, Delic-Dayal: A New Analytics Perspective, pp. 26-32, Vol. 6, No. 11, June 2003

[2] Active Enterprise Analytics, HP OVUA Conference, Nice, Summer 2006.

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s