Thinking Cloud Analytics
Analytics providing insight into applications moved to the cloud, the user experience, and usage patterns can be used to inform business decisions.
You will be in the cloud -- if not now, then soon. You have tools for analyzing your on-premises systems, and you'll need tools to analyze your cloud implmentations, too. These tools should provide the analytics necessary to make informed business decisions.The Cloud Decision
Many CIOs are concerned with a number of questions relating to the cloud, and some of these questions enter the decision-making process itself: How does the CIO determine which applications should be moved to the cloud? How does the CIO define the desired application architecture? What capacity and resources are needed? How can we ensure the end-user experience will be as good if not better than today? Will it be costly to make the move? ... Here is where it gets to be interesting: How does the CIO and staff monitor progress and measure success?
In-depth analytics providing insight into applications moved to the cloud, the user experience, and usage patterns can help address all of the above questions.
I reached out to one such analytics tool, SoftWatch, to learn more about cloud analytics for supporting cloud migration, and co-CEOs Uri Arad and Moshe Kozlovski armed me with information. SoftWatch is a leader in application usage analytics. With its SaaS solution, it enables enterprises to effectively manage the transition of business applications to the cloud, optimize their hybrid cloud environment, and reduce software spending.Application Profiling
Application profiling is based on analyzing the resources consumed when supporting an application. These resources include the CPU, disk I/O, system memory utilization, and network utilization. The profile of the end-user provides information about the actual usage of the application. This can include segmentation of users to occasional, medium, and heavy users and their usage patterns. The CIO can compare the application behavior and deliver metrics suitable to be presented to the decision-makers.Do Analytics Vary for IaaS, PaaS, and SaaS?
There are similarities such as possible degradation of productivity, user resistance to change, poor change management practices, etc. Having said that, it is expected that core applications that are not "off the shelf" products may involve additional risks that are related to technical aspects. These include architectural design, performance, capacity planning, and security that tend to be more complicated in many cases than in adopting a SaaS solution.What Can Be Measured?
We measure end user usage of the application and application re-source consumption. The end user activity includes inactive users, read only users, light users, and power users. The resource consump-tion includes utilization of CPU, I/O, memory, and network throughput. [Some of these measurements can be used to validate the cloud us-age billing.]Why is it Important to Relate Resource Usage to End User Behavior?
In order to make informed decisions, a comprehensive view of the application usage is essential. Examples are:
- Planning the needed capacity is related to resource consumption and identifying peaks users' activities
- Making the decision whether to migrate to a cloud platform vs. an available SaaS solution such as UCaaS
- Determining the architecture considerations and the costs that are affected by the consumption of physical resources
Managing a hybrid IT environment of on-premises applications and cloud applications is rather complicated and requires ongoing under-standing of resource consumption and user activity that is provided by the metrics. This serves ongoing decisions and operational activities such as:
- Selecting the next application to move to the cloud
- Planning and executing training programs
- Adjusting licensing models to support the hybrid environment
- Reducing costs
- Monitoring the progress of adoption of the new cloud application
- Applying change management tools to expedite the transition
In most cases, ongoing resource allocations will be supported by the tools provided by the cloud platform vendors. However, this is ra-ther expensive and the ability to forecast needed resources that are related to user behavior (peak periods for example) can help in fore-casting an accurate resource consumption and capacity needs and in negotiating a more predictable, less costly resource allocation.Can This be Applied to Communications and Collaboration Cloud Services?
SoftWatch provides full monitoring of email and Web conferencing tools. As to collaboration, SoftWatch can identify collaboration patterns in Microsoft Office applications and this serves in planning the transi-tion to collaboration cloud services (such as Google Apps and Office 365).What Can the Enterprise do with the Metrics Cap-tured?
Support major initiatives including:
- Move to cloud-based applications
- Move to cloud infrastructure
- Optimize license spending for desktop applications and Web ap-plications
- Optimize end user computing environment
- Standardize cloud storage and Web conferencing tools
- Monitor and manage the adoption of cloud services