Our classical model is the airplane 'black-box', or data recorder. This has evolved over more than 50 years, and is a popularly accepted and dependable technology to resolve airplane crashes by its data alone. In the computer industry, the model was followed and continued, 40 years ago, by mainframe platforms ("operating systems"). Other tools can facilitate and often provide necessary, but not always sufficient information to resolve a computer problem on its first occurence.

 

The difference can be found in problem detection tools (let you know if a server is 'down') and problem matching/'screening' tools to match the symptoms of a current problem to the symptom and fix of a prior problem vs. detailed internal data to solve a new and unique problem on its first occurrence. Problem detection tools and screening tools are valuable, but what is really most valuable are tools that enable root-cause problem-solving. We assist in the use of pre-existing tools,  and creation of new tools, with experience designing and using both kinds of tools. In addition, we will guide systems design for first-fault serviceability capability.

 

See First Fault Tools  for a fuller description of the capabilities of these tools. The list of products that 'monitor' a computer system for performance and/or reliability faults is very large - there are many such products. For workload imbalances, these tools are sufficient. But for complex internal problems, there are very few root-cause-solving products. We have a wide range of vendor products to point to for instances of these capabililties, and we have direct experience with many of them.

We provide advice, via consulting engagements. Our advice will facilitate understanding a particular organization's services and/or products, and making improvements to their current attainment. We will help organizations further exploit technology already present at their site, make investments in additional products, and help them develop improved first-fault problem resolution features and products, as each organization feels appropriate. We can benefit both end-user 'data-center' or system deployment organizations, and also software product developers - both types of organizations can benefit from improvement in their First Fault Problem Resolution capabilities. Our consulting engagements are customized, of course, but can include basic assessments such as first-fault problem resolution success (how many problems are resolved directly, vs. how many problems neeed further reproduction and additional 'instrumentation'). We can and have done "Availability studies" to find opportunities for improvement - these necessarily point to the value of improving speed and accuracy of that initial fault causing the "availability" problem.