A presentation at Devopsdays London 2019 in in London, UK by Nayana Shetty
Monitoring Decoded Why, What and How? Nayana Shetty Tech Lead, Infrastructure Delivery Financial Times @shettyny
@shettyny
Proactive Issue Detection @shettyny
Quick & Easy Incident Management @shettyny
Happy Service == Happy Customers @shettyny
@shettyny
Checklist of predicted problems @shettyny
Use Templates for Monitoring @shettyny
USE METHOD, by Brendan Gregg @shettyny
RED METHOD, by Tom Wilkie @shettyny
RED method is about how happy our users are while USE method is about how happy our machines are. by Tom Wilkie @shettyny
@shettyny Picture by Denise Yu(@deniseyu21)
@shettyny
Monitoring Tools @shettyny
Monitoring capabilities @shettyny
Log Aggregation @shettyny
Log Aggregation Metrics Aggregation @shettyny
Heimdall Visualisation Log Aggregation Metrics Aggregation @shettyny
Heimdall Visualisation Log Aggregation Metrics Aggregation Basic Healthcheck __health @shettyny
Heimdall Visualisation Log Aggregation Alerting Metrics Aggregation Basic Healthcheck __health @shettyny
@shettyny
Some of the common question engineers have around monitoring are, “Why should I monitor my service?”, “How do I monitor my service?” and “What tools do I use to do this?”. In this talk we will go through some of the key reasons for us to monitor our services and systems. We will then see some methods like the USE(Utilisation, Saturation and Error) and RED(Rate Error Duration) methods that help us identify what to monitor in a consistent manner. Lastly, we will cover some broad classification of tools that help us achieve this including some real examples of monitoring some systems and services at the Financial Times.
for free. You
can too.