5 Myths About DevOps Busted


DevOps as a service complements the Agile Software Methodology by promoting the communication and collaboration between Development, Quality Assurance and Operations, thereby enhancing the IT Organization’s Capabilities in accommodating rapid changes to Production and Remediating Production Issues as they occur.

But, most of the time, the Core Principle about DevOps is not understood properly. Some of general Misconception/Myth that are associated with DevOps are:

1. Operations + Development = DevOps:

Combining team members from Operation and Development can establish a DevOps operation. This is just a myth. The fact is DevOps is a combination of Processes and Practices that are adopted for the entire delivery pipeline and covers various stakeholders, where two key practices are adopted such as Continues Integration (CI) and Continues Delivery (CD).

2. Complete Automation is DevOps:

DevOps is all about Completely Automating the Build and Release Process. Automation in DevOps forms an integral part of the process, but implementation of Automation is not the only activity associated with DevOps. In DevOps, Process Automation is achieved by using tools such as Puppet, Chef, Ansible, Etc., which are available in the market. It has to be noted that Implementation of Automation should be limited to an extent where it remains under control. Implementation of Complex Automation Scripts in Complex Server can simply end up as a road block, rather than being a solution.

3. DevOps is a Tool:

Implementation of Configuration Management Tools is DevOps. Actually, DevOps has nothing to do with tools implementation. It is just part of the process and as said earlier it helps in Process Automation. Experts often feel that tools in reality undermine the DevOps potential. Automation and Tools form only part of DevOps and the core remains with combining and increasing end-to-end practices of collaboration with CI or CD.

4. No More Traditional IT Roles with DevOps

Each and every role needs to be handled by all the members of a DevOps Team Eliminating any Individual Roles. The real objective of DevOps is to Eliminate the collaboration barrier and not to ask everyone to work by adopting to roles. To make the Support Operation Effective, Specialized Skills and Traditional Roles are valuable for DevOps.

5. DevOps needs a Dedicated Team:

DevOps requires a dedicated DevOps team. This is not true and DevOps gives more importance to the processes rather than focusing on a dedicated team or dedicated role. There are some occasions when the mission of the DevOps team in not defined properly and having a dedicated team leads to more problems. There are even occasions when a temporary DevOps team makes more sense to help streamlining processes.

The fact remains that DevOps as a concept is still evolving and it is bound to get buried under myths. All efforts should be directed towards bursting these misconceptions and align organizational goals with the DevOps principles to maximise ROI.

KARYA’s DevOps Offerings:

Our versatile team of DevOps professionals include top-notch Architects, Automation & Integration Specialists, Application Architects, Open Source Technologists, Cloud Orchestrators, System & Database Administrators, Quality Assurance Professionals, Build / Release / Deployment Managers can help you improve Code Quality, Integrate Continuously, and Deliver Faster. To find out more about KARYA’s approach to achieve Continuous Software Delivery CLICK HERE.

(About the Author:- Praveen Kumar Rajendran works as a Senior Consultant- Presales at KARYA Technologies. He holds Masters’ Degree in International Business from La Trobe University and has wide experience in Business Consulting on various technologies. He loves to write on latest trends in IT and his areas of interest include Mobility, Cloud and Enterprise Solutions.)

How To Maintain Homeostasis Using Business Intelligence Alerts

How to maintain Homeostasis using Business Intelligence Alerts

What is Homeostasis?

Encyclopaedia Britannica defines Homeostasis as any self-regulating process by which Biological Systems tend to maintain stability while adjusting to conditions that are optimal for survival.

If Homeostasis is successful, life continues; if unsuccessful, disaster or death ensues. The stability attained is actually a dynamic equilibrium, in which continuous change occurs yet relatively uniform conditions prevail. An often cited example for Homeostasis is temperature regulation in humans and warm-blooded animals in general.

Key Components

The two important components of Homeostasis are Sensors and Effectors. Sensors monitor and detect changes in the controlled entity and provide the negative feedback for Effectors to carryout corrective measures and these corrective measures remain in motion until the deviation is reversed.

The effectiveness of Homeostasis is in its ability to trigger the corrective measure in real time while the controlled entity is alive, well and able to survive the deviation and also self-correct. It is borne out of survival instincts, the need to survive and perpetuate.

An organization is a controlled entity and in order to effectively monitor and regulate the various processes within it, it is worth investing in the Sensor and Effector Mechanism because they ensure its survival. Homeostasis is nothing new in the business world. Most manufacturing processes are monitored using Industrial and Engineering Control Systems that are self-regulating.

In IT Services the advent of DevOps is a direction towards a seamless response system for dynamic demand on Software, Services and Infrastructure. Machine Learning, Algorithmic Processing of events and transactions and Internet of Things are good examples of embedded Homeostasis in action. In the end, a living optimal entity is a result of an ever present Homeostasis.

Alerts and Sensors

Organizations can leverage their existing Business Intelligence Infrastructure to install Alerts. There is lot more emphasis on the need for On-demand Decentralized Reporting than demand for Auto-course Correcting Mechanisms that BI Alerts can facilitate. BI Alerts are the Sensors. Alerts trigger when certain pre-determined set points are deviated and provide the negative feedback necessary for the Effectors to kick in. These set points are commonly called KPIs but I would like to bring in measures for vitality instead of KPIs. I think Vitality and Homeostasis go together.

Many organizations use Alerts and Notifications from their BI Systems, but not as Cohesive Mechanism for an Enterprise-wide Alarm System to stabilize operations and survive. Balanced Scorecards measure and monitor performance but do not help in keeping Homeostasis in the earliest available opportunity. In order to get to the deep green colour highlight on a Balanced Scorecard, Alerts are a must for course correction.

What must you sustain on a daily, weekly or a monthly basis to remain vital. If the answer is based on Average Foot Fall or Customer Complaints or Sales or Margins. A very simple example of Vitals calculated from these are;


Vitality of Sales = Inclination of Trend line of (Weighted Average YTD Actual Sales (Daily)) minus (Weighted Average YTD Actual/Budgeted Sales (Yesterday/Daily))
{Trigger Alert when Vitality of Sales < 0 degrees} {Stop Alert when Vitality of Sales >= 0 degrees}

Customer Complaints

Vitality of Customer Satisfaction = A Trend line of (Average YTWE NPS (this week) – Average YTWB NPS (week before))
{Trigger Alert when Vitality of Customer Satisfaction < 0 degrees} {Stop Alert when Vitality of Customer Satisfaction >= 0 degrees}

The Trends and Averages may be daily, weekly, monthly or some other period depending on how volatile your parameters are. The average chosen should iron out daily fluctuations and clearly show a trend upward, downward or steady state. Instead of a plain vanilla Average Calculation, a sophisticated Forecasting or other Statistical Models relevant to your industry may be used to calculate the trend and define the Alert set points.

Predictive Analytics to forecast a set point or a trend and thereupon issue an alert can be an effective red flag. Alerts may be graded to denote the significance of the deviation. Since we are talking about Homeostasis, we may also borrow the concept of Early Warning Score (EWS) used by the Medical Community to assess the degree of illness of a patient. In BI Terminology we know it as tolerance threshold in a Balanced Scorecard. Preventive Action kicks in when the Alert goes out in response to meeting an EWS.

‘The Alert System’

The Effectors are the organizational managers and their teams, who have the resources to diagnose the reasons for deviation from Vitals, take corrective measures. On-demand Reporting can be a great tool for Diagnosis. The alerts must be designed to facilitate Drill-through Action to deeper levels of data.

In fact, “Project Homeostasis”, cannot be a success without a robust Data Processing and Integration Infrastructure in place for Analytics. The beauty and the effectiveness of ‘The Alert System’ is in its ability to keep blaring its horns until the fire has been doused. There are no accessible buttons to push to silence the alarms. They can only be silenced by correcting the course of the Trend Line.

Knowledge about your Business, Industry in general and a Creative Imagination are the limits to how Alerts may be defined and implemented in your organization.

KARYA can help organizations design Enterprise-wise Alerts by leveraging your existing Business Intelligence Systems. To know more about KARYA’s Data Management Solutions please log on to http://www.karyatech.com.

(About the Author- Nagarajan Mahadevan works as Principal Consultant- Data Management Solutions at KARYA Technologies. He is a Techno-Commercial expert and has more than 2 decades of experience in Business Intelligence, Data Warehousing Technologies, Financial Accounting and Management Consulting. His areas of interests, apart from staying abreast on latest IT Trends and Technologies, are Yoga and Indian Classical Music.)