Proactive Monitoring: Playing Offense for the Win
2 likes•498 views
The document discusses proactive monitoring in IT, highlighting the limitations of current reactive methods that rely on thresholds and metrics. It emphasizes the potential of AI to predict issues before they affect business operations through three types of predictions: behavioral, probabilistic, and semantic. The presentation outlines a shift from traditional monitoring dashboards to AI-driven solutions that can autonomously monitor and predict system failures.
Proactive Monitoring: Playing Offense for the Win
- 1. Proactive Monitoring: Playing Offense for the Win Dror Mann | CO-Founder, VP Product
- 2. Confidential and Proprietary Dror Mann VP Product & Co-Founder Loom Systems Head of Product Voyager Labs Head of Operations IDF elite Intelligence unit [email protected] Hi!
- 3. Confidential and Proprietary • Intro – The Digital Era • Today’s Monitoring • What can we expect from AI • Proactive Monitoring, 3 types of Predictions • Live Demo • Q&A Agenda
- 4. Reality
- 6. Confidential and Proprietary We get you!
- 7. Confidential and Proprietary Detect issues using Metrics (Thresholds) Monitoring Today
- 8. Confidential and Proprietary Monitoring Today Drill down using Logs
- 9. Confidential and Proprietary • Reactive – Detecting issues after they happened - see #oncallselfie • Extracting metrics and maintaining thresholds takes time • Metrics tell you (only) part of the story • You monitor (only) what you know Problems
- 10. What if there’s a better way?
- 11. Confidential and Proprietary Pattern Recognition Strict Methodology Large Dimensionality HUMANS Good at top-down, open-ended tasks BOTS Superior at defined, bottom-up tasks Deep reasoning Contextual thinking (Hungry) (Get tired)
- 12. Confidential and Proprietary • Proactive – Tell me about issues before they affect my business • Monitor everything! Detect Unknowns and Knowns • Automatic as possible - without defining thresholds • Yes, also logs, they have better answers: Desired Solution Credit:XKCD
- 13. Confidential and Proprietary “I’ve been hearing this for 20 years” Total Recall, a movie based on a book from 1966, featuring a self-driving car as science fiction. If Artificial-Intelligence has matured enough to drive your car, it can probably also help with your IT. Skeptic?!
- 14. Collect + Parse Learn Baselines and Predict failures Correlate & Trace Enrich with Insight/Action Proactive Monitoring using AI
- 15. Confidential and Proprietary • Behavioral • Probabilistic • Semantic 3 Types of Predictions
- 16. Confidential and Proprietary Behavioral Vector analysis Rate of change Threshold
- 18. Confidential and Proprietary Semantic Certificate for local system with Thumbprint 7e 2f ce f9 7e 33 fb 1a 59 16 f5 0f a5 d3 fc a2 ed 18 21 2f is about to expire or already expired. Source: Microsoft-Windows-CertificateServicesClient-AutoEnrollment Date: 22/06/2012 2:43:56 PM Data Meta Data 2017-07-21 04:23:35 app web.1 WARN : Detected yajl-ruby version 1.1.0 which can cause segfaults with newrelic_rpm's thread profiling features. We strongly recommend you upgrade to the latest yajl-ruby version available. 2017-07-21 04:22:17 app web.1 WARN : Detected yajl-ruby version 1.1.0 which can cause segfaults with newrelic_rpm's thread profiling features. We strongly recommend you upgrade to the latest yajl-ruby version available. 2017-07-21 04:21:17 app web.1 WARN : Detected yajl-ruby version 1.1.0 which can cause segfaults with newrelic_rpm's thread profiling features. We strongly recommend you upgrade to the latest yajl-ruby version available.
- 19. Poll Question
- 20. Let’s see it live
- 21. Confidential and Proprietary Past vs. Future Past: ITUSED DASHBOARDS TO DO MONITORING Future: AIDOES THE MONITORINGFOR IT
- 22. Q&A