We’ve heard it time and time again: one of the biggest challenges for Hadoop operators is too much time spent troubleshooting. Whether it takes way too long to put out figurative fires on your Hadoop cluster, or you spend forever trying to determine the root cause of performance issues, or you can’t for the life of you work out why jobs are running slowly – troubleshooting is a constant drain on time and business value. Today’s Hadoop users need the kind of advanced troubleshooting assistance that comes from deep introspection into hardware usage at the user, job and task/container level. Unfortunately, most monitoring tools (like Ganglia, Ambari, Cloudera Manager, etc.) fall short because they only provide performance metrics at the node level.
Enter Pepperdata Insights Service, the most comprehensive Hadoop troubleshooting technology available today, rolled into a simple, easily-deployed one month program. With Pepperdata Insights, organizations can use the Pepperdata software for one month – with no limits on cluster size – and experience a 90% reduction in troubleshooting time! It’s the first solution that addresses the shortcomings of today’s tools and helps Hadoop operators troubleshoot performance problems on their clusters with minimal time and effort. With Pepperdata Insights, you can get hours – even days – of time normally spent troubleshooting back because the software provides both a macro and granular view of everything that’s happening across the cluster, making it easy for users to drill down into root causes without having to rummage through logs and/or configurations.
Pepperdata Insights provides detailed performance metrics that allow users to break down overall usage by queue, group, and user – and then drill down to individual jobs and containers so that when something goes wrong, you are able to identify what, where, and when it happened, instantly. On top of that, at the end of the 30 day subscription, organizations receive a full diagnostic report of their cluster activity, accompanied by expert recommendations from the Pepperdata team on how to improve performance and guarantee service levels moving forward. It’s the closest you can get to a physical for your cluster by seeing how the overall system is working, with specific recommendations on how to improve the “health” of your operations.
