How to be a Monitoring Consultant
summary
Recommendation: focus on the pragmatic recipe — start with top ten systems not auto-discovery, prefer state change checks over noisy time-series for operational alerts, build custom plugins (example: a triple‑redundant TV data check that enforces 1‑minute accuracy), automate from a single source of truth using NetBox + Ansible integration, avoid brittle SNMP, design an explicit alert lifecycle (soft/hard states, notify/escalate timings), and prefer distributed checks and event commands only as temporary self‑healing, not a fix for bad design.
AI-written summary · about
chapters (16)
- 0:00Introduction and background
- 3:37What monitoring means to people
- 8:12Monitoring as service delivery
- 10:44Starting with top priorities
- 14:58Lifecycle of monitored events
- 16:46Defining business priorities
- 19:36Choosing the right tools
- 23:41Writing custom plugins
- 26:54Automation and monitoring nirvana
- 27:56Demo of NetBox IPAM
- 31:52Event commands and automation
- 33:16Metrics and alert enrichment
- 35:01Monitoring in project planning
- 36:47Common pitfalls and firewalls
- 38:15Design and sustainability
- 39:59Q&A session
related talks
18mThe Good, the Bad and the Ugly: The 3 Learnings of an SRE
39mConfessions of a Systems Engineer: Learning from My 20+ Years of Failure
39mThe Beginning, Present, and Future of Sysadmins
20mScale up your job satisfaction, not your software
19mThe Never-Ending Story of Site Reliability
30mWhere Do I Go From Here?
50mSafe Landings: Self-care In A High Intensity Technological World
29mManagement is hard job! That's why you should do it like an engineer