Hotspot Question You have an Azure Load Balancer named LB1 that balances requests to five Azure virtual machines. You need to develop a monitoring solution for LB1. The solution must generate an…

Question

Exhibit

Answer Area

• An unavailable virtual machine:Data Path Availability
  Byte CountData Path AvailabilityHealth Probe StatusPacket CountSYN Count
• More than 50,000 connection attempts per minute:SYN Count
  Byte CountData Path AvailabilityHealth Probe StatusPacket CountSYN Count

Explanation

This hotspot question tests your knowledge of Azure Load Balancer metrics and which specific signals/metrics to use when configuring alerts for availability and connection rate conditions.

Approach. For detecting an unavailable virtual machine, you should use the 'Health Probe Status' metric (also referred to as 'Data Path Availability' or 'Health Probe Status'), which reports whether backend instances are responding to health probes - a value below 100% indicates one or more VMs are unavailable. For detecting when connection attempts exceed 50,000 per minute, you should use the 'SYN Count' metric, which tracks the number of TCP SYN packets (new connection attempts) sent to the load balancer, making it the correct signal to alert on high connection attempt rates.

Concept tested. Azure Load Balancer monitoring and alerting - specifically understanding the difference between the Health Probe Status metric (used to detect backend VM availability) and the SYN Count metric (used to track TCP connection attempt rates). These are native Azure Monitor metrics available on Azure Load Balancer (Standard SKU) that can be used to create metric-based alert rules in Azure Monitor.

Reference. https://docs.microsoft.com/en-us/azure/load-balancer/load-balancer-standard-diagnostics

No community discussion yet for this question.