1 PM EDT
The Solarwinds Monitoring Cloud companies are experts in how to do monitoring at scale in very high throughput environments. During our time running on the cloud, we’ve used many DevOps concepts, including automation, to build a fully-instrumented stack from infrastructure all the way through the application that allows us to have observability into all areas of the stack and from multiple angles.
Professional Resume Template for Software Developers
Monitoring First - Instrumenting Your Entire Stack for the Ultimate in Observability
1. Monitoring First: Instrumenting Your
Entire Stack for the Ultimate in
Observability
Dave Mangot
Head of Site Reliability Engineering,
SolarWinds Cloud Companies
@davemangot
8. MDD
• Instrumentation-as-Code
• Single Source of Truth
7
By Minesweeper (Own work) [CC BY-SA 4.0 (http://creativecommons.org/licenses/by-sa/4.0)], via Wikimedia Commons
11. MDD
• Instrumentation-as-Code
• Single Source of Truth
• Developers Curate Visualizations and Alerts
• Alert on What You See
• Don’t Measure Everything
10
12. "Just because you can measure
everything doesn't mean that you should."
—W. Edwards Deming
11
13. MDD
• Instrumentation-as-Code
• Single Source of Truth
• Developers Curate Visualizations and Alerts
• Alert on What You See
• Don’t Measure Everything
• Show me the Graph
12
16. Nagios, etc.
Is a port listening?
Is that what I really care about?
15
17. Nagios, etc.
Is a port listening?
Is that what I really care about?
Are the number of requests over a period of time below a threshold?
What are my average # of requests?
What are the min/max/std dev of latency?
16
19. 2nd Way of DevOps
The Second Way is about creating the right to left feedback loops. The goal of almost any process
improvement initiative is to shorten and amplify feedback loops so necessary corrections can be continually
made.
The outcomes of the Second Way include understanding and responding to all customers, internal and
external, shortening and amplifying all feedback loops, and embedding knowledge where we need it.
18http://itrevolution.com/the-three-ways-principles-underpinning-devops/
34. Application Instrumentation
33
# Timing a potentially latent operation with custom instrumentation
statsd.time("api.measures.source_match_build.time") do
metric_source_match_build(active, explicit, wildcards, negates)
end