The intersection of machine data analytics and the internet of things

SRIKANTH DESIKAN
Sunday, September 8, 2013

The ability to gather and harness data from machines and devices connected to the net is increasingly becoming a source of competitive advantage for those who have been thinking ahead. GE is widely cited, with their INDUSTRIAL INTERNET initiative. Cisco has been talking about INTERNET OF EVERYTHING for a while. At Glassbeam, we have been focusing on the infrastructure required to process and analyze machine data for over 3 years now. Large companies like IBM and Hitachi have benefitted from digging deeper into their data. So far we have processed over 100,000 devices, hundreds of millions of events and sections and terabytes of multi-structured log data. This experience has set us up to analyze any data from any machine. While there are several good log processing options for simple data center logs and weblogs, complex machine data requires a more SOPHISTICATED APPROACH to parsing, transforming and loading the data. Further the benefits are to business users and hence the application exposed cannot be complex, requiring them to write esoteric search expressions to see reports and analysis. Glassbeam simplifies the app experience for the business user, thus driving very quick ROI for business looking to gain value from their machine logs.

Machine data can help answer so many questions about your products and applications. There are many ways our customers use Glassbeam – here are some key questions and queries resolved by Glassbeam

  1. Which customers are at risk for an upgrade because of a known bug in a firmware?
  2. How many versions of our products are being used worldwide in our installed base?
  3. Which customers are running in what platforms and OS?
  4. What features of our products are being used by which customers and customer segments?
  5. How can I detect changes to configuration on my customer’s machines that can cause a spike in support issues?
  6. How can I identify patterns of usage or configuration that are most correlated to customer satisfaction issues.
  7. How do I detect the root cause of a problem by correlating errors and issues in some of my log files with changes to the system environment or settings?
  8. How can determine how many procedures were performed by doctors at various settings on my MRI machine by hospital by doctor.
  9. How can search for known patterns in my logs periodically to solve specific problems at my customer site or application?
  10. How do I know when my products are going fail or when a part needs to be replaced pro-actively since replacing a part before its due costs a huge amount of money?
  11. How can I get a birds eye view of storage capacity across all of my sites and get alerted when a threshold is breached?
  12. How do graphically trend various metrics captured in my logs and determine correlations between application performance and configuration of my VM, Network , Storage system or my servers.
  13. When I troubleshoot with my customer’s log, I want to apply a database of previously created rules and determine if there is close match for any known issues with that log.
  14. How many products are in the field with unsupported versions or components that are causing us unnecessary spend on support and maintenance.

There are hundreds more but thats a sample. Glassbeam’s apps helps your arrive at answers to such questions based on your machine data and if needed other structured data. If you are grappling with such issues would love to hear from you.