Links

• AWS Control Tower: For managing a multiple accounts strategy consistently with automation. (Traceability, compliance, etc)
  • 
• In which honeycomb talks about their decision to move certain services to kubernetes
• Full stack java development with spring boot and vuejs
• A few things to keep in mind about where dashboards are good tools and where they come up short:
  • I guess they represent how the creator thinks about the important aspects of a system at a single point in time (keeping them up to date in quickly evolving systems is a challenge)
  • Correlation between metrics and a particular can be hard. Sometimes an incident impacts many metrics
  • Having to analyze lots of graphs that may or may not be relevant to an investigation is hard work for someone who is well rested, during daylight hours. If one is responding tired, at night the analysis becomes harder
  • People add metrics to dashboards for incidents until they become large and unwieldy
  • Are useful for well known, high level metrics eg KPIs, or a patient hooked up to a diagnostic machine where vitals are being monitored (blood pressure, heart rate). Used this signal as a way to get a sense that something isn’t right but dig deeper with other tools