Working with our customers on analyzing their machine log data, we found various common usage patterns of the log data. First, the users in a support organization are troubleshooting and hence are most interested in searching, viewing logs and looking at specific sections or exploring various logs in a log bundle. This is a basic use case that requires correlating various logs in a bundle, whether they have a time stamp or not and allowing users to do full text and attribute based search on the logs( attributes are the output of parsed logs).
The next usage pattern is from level 3 engineers, development and engineering organizations or even product managers looking to “explore” their log data over a time period looking for patterns across customers or plotting trends for a specific parsed attribute over time etc. The Glassbeam search application is flexible and business-user friendly – no complicated queries and search language is needed and users can view and explore their data using “faceted” search and Quickgraphs.
Now the same set of parsed data is extremely valuable for root cause analysis, customer facing dashboards and deeper product and installed-base analytics. The Glassbeam dashboards provide a business-user centric UI that can be customized based on roles and users, providing a complete 360 degree view into the installed base and customer usage.
Finally, data science teams at our customer sites, now can leverage the structured data created from logs by the Glassbeam platform to run models and predict machine failures, part replacement times etc. This complete end to end value from machine data provides a very high ROI to our customers since the entire enterprise benefits from the information in their machine log data.
Multi-function benefits from machine log data