Implement an instrumentation strategy Flashcards
What percent of the total exam questions are dedicated to this subject?
10% to 15%
What is an instrumentation strategy?
A strategy which when implemented provides logging, telemetry, and monitoring to give you full visibility into what’s happening in your environment.
What are the main tools for monitoring a DevOps environment in Azure?
- Agent based (install agent or SDK to exsiting agent e.g. application insights or new relic)
- Agentless ( some tools don’t require installation or modification e.g. application insights has agentless features)
- Microservices (e.g prometheus or azure monitor for containers)
- Integrated platforms (monitor the infrastructure your application is running on like hardware or OS e.g. azure monitor)
What does APM stand for?
Application Performance Monitoring
What are the steps of a platform monitoring feedback loop?
The feedback loop consists of Plan -> Build -> CI -> Deploy -> Operate -> Continuous Feedback
The following help achieve that:
- Enable monitoring for all applications and infrastructure
- Combine resources in Azure Resource Groups
- Ensure quality through continuous deployment
- Create actionable alerts with actions
- Create dashboards and workbooks
- Continuously optimize
How to ensure the safety of the data generated by a monitoring platform?
- Use built in monitoring roles
- Security considerations for monitoring data
- Monitor within a secured virtual network
How can pipeline events be used within an effective monitoring platform?
PIpeline events
What is an action group?
A collection of notification preferences defined by the owner of an Azure subscription
Azure Monitor and Service Health use action groups to trigger alerts. What is the fundamental requirement to alert alert users?
An action group, which is a collection of notification preferences defined by the owner of an Azure subscription.
How many action groups can be defined in a single subscription?
s
Up to 2000.
What are the steps to execute logic based on monitoring results?
Alert -> Action group -> Logic App
What is distributed tracing?
When the application is more complex than a single web app talking to a single database, distibuted tracing can be used to monitor complex and layered calls accross a distributed system.
This is necessary in modern micro services architecture or distributed computing.
How to enable distributed tracing?
- Enable via Application Insights through auto-instrumentation or DKS’s
- Enable via OpenTelemetry
- Enable via OpenCensus
How can you detect performance anomalies and get help understanding what users do with your app?
Application Insights detects performance anomalies. It has agents running in the various scopes of your application (runtime, build time and client side) and the information gatherd is brought together to help you get a birds eye view of whats happening accross your stack.
What are the broad sources of logs and metrics?
- Application
- OS
- Azure Resources
- Azure Subscription
- Azure Tenant
- custom sources.
What is a log analytics workspace?
Logs and metrics usually pass through Azure data processing services to get insights, visualize them etc, after which they are pushed to the underlying storage, a log analytics workspace.
What are the recommended metrics for ensuring sucess?
- Success vs Failure
- Timing Metrics
- Throughpout Metrics
What are the reccomended best practices to integrate user analytics in an Azure application?
- Send telemetry from your app with application insights
- Design telemetry with the app, developing the code with the telemetry in mind as you do it
- Track custom business events with thoughtful logging during execution
- Explore usage demographics and statistics with user sessions report filter
What are kusto queries?
Kusto queries are used to query logs
What does the kusto verb render do?
Render is a part of the visualisation features of Kusto. It can render results are various visualisations
