HPE Software is now Micro Focus
HPE Software is now Micro Focus
IT Operations Management (ITOM)

Inside a new approach to analyzing machine log data

Inside a new approach to analyzing machine log data


Operations Analysis IT Operations.jpgGuest post by Sancha Norris

Sr. Product Marketing Manager

Operations Analytics


How do you solve a performance problem when the cause is unknown, you don’t know where to start—and time is of the essence?


If you work in IT Operations, this is one of those troubling questions that keep you awake at night. It’s right up there with pesky intermittent performance problems that impact performance and availability. Because they come and go, they’re nearly impossible to catch, identify their root causes and fix them.


The answer to both scenarios lies in machine data. Today’s IT environments contain a wealth of information about system usage, performance, events, configuration changes, customer data, business metrics and so on. Increased efficiencies and capacities of modern storage systems have reduced the cost per record for data storage and made it financially practical to capture, store, and analyze more log data over longer periods of time.


Download the new HP white paper

“Analyzing Machine Data—The Best Way Forward”


Of needles and haystacks

However, while logs have played a role in troubleshooting and raw data analysis, the overwhelming growth in log data in recent years—and the unstructured nature of that data—has made it nearly impossible to find that crucial insight, “the needle in the haystack,” that can accelerate the remediation of the problem and prevent future problems. Analyzing massive volumes of logs can still be a very manual process, and traditional log search techniques are simply not keeping pace.


It’s little wonder that as potentially valuable a source as machine log data could be, most companies have not yet fully and effectively utilized it to troubleshoot performance problems.


New approach to log data analysis


Now a new white paper, “Analyzing machine data—the best way forward” describes a more effective approach to harvesting the insights that are hidden within log data. The approach is based on automating log analysis and applying sophisticated machine learning to the operations analysis process, so that you can pinpoint the root cause of performance issues in minutes rather than in hours or days.


The white paper introduces a more intelligent approach to log analysis and examines three critical aspects:


  1. Group similar logs together to expedite processing
  2. Use machine learning to find patterns and determine relevance of logs
  3. Tune log analytics with SME expertise to optimize accuracy


Download the white paper to learn more about how you can extract highly relevant, actionable insights from your log data using automated, systematic analysis to quickly identify the root cause of performance issues.


Download “Analyzing Machine Data—The Best Way Forward”


Learn more


Watch these videos about analyzing machine data with HP Operations Analytics and how you to use Log Analytics to help you find the root cause of an IT problem. Download datasheet here.


Visit hp.com/go/OpsAnalytics

  • operational intelligence
0 Kudos
About the Author





I am from TCS and my job is to troubleshoot high severity performance problems across all TCS accounts. I  run the internal SWAT team to help the accounts in time of crisis. The key issue is, we need to identify the Root Cause pretty fast, so I can relate very well to the "needle in a haystack" paradigm.

We do use a lot of analytics on both structured as well as unstructured logs as well as monitoring tool/ event management tool data. Normally we use Splunk, which we find extremely handy. While machne learning is not supported by Splunk, its not our use case as well. We normally jump into unknown environments to troubleshoot Sev 1 issues !

Would love to evaluate the tool! Pl let me know how I can do it.




Sarthak Banerjee

Chief Architect, Delivery Governance



Honored Contributor.

Hi Sarthak, we would be happy to show you Operations Analytics. Would you send me a quick email and I'll get you connected to the right person?