In the context of AWS Load Balancers, the term “threshold value” can refer to several parameters or settings used to manage and monitor the performance and health of the load balancer and its targets. Here are some key threshold values commonly associated with AWS Load Balancers, specifically Application Load Balancers (ALBs) and Network Load Balancers (NLBs):
1. Health Check Thresholds:
- Unhealthy threshold count
- Healthy threshold count
- Health Check interval
- Health Check timeout
1. Unhealthy Threshold Count: The number of consecutive health check failures required before considering a target as unhealthy.
- Default value: 2
- Range: 2-10
2. Healthy Threshold Count: The number of consecutive health check successes required before considering a target as healthy.
- Default value: 5
- Range: 2-10
3. Health Check Interval: The time between health checks of an individual target.
- Default value: 30 seconds
- Range: 5-300 seconds (5 minutes)
4. Health Check Timeout: The amount of time, in seconds, during which no response means a failed health check.
- Default value: 5 seconds
- Range: 2-120 seconds
2. Connection and Load Management Thresholds:
- Connection Timeout
- Idle Timeout
1. Connection Timeout: The maximum time in seconds that the load balancer waits to establish a connection to a target.
- Default value: 60 seconds
- Range: 1-3600 seconds (1 hour)
2. Idle Timeout: The amount of time that a connection is allowed to be idle before being closed by the load balancer.
- Default value: 60 seconds for ALBs
- Range: 1-4000 seconds (approximately 1.1 hours)
3. Target Group Attributes:
- Deregistration Delay
- Slow Start Duration
1. Deregistration Delay: The amount of time to wait before deregistering a target during a scale-in event. This delay allows in-flight requests to complete.
- Default value: 300 seconds (5 minutes)
- Range: 0-3600 seconds (1 hour)
2. Slow Start Duration: The time period, in seconds, during which a newly registered target receives a linearly increasing share of the traffic to allow it to warm up.
- Default value: 0 seconds (disabled)
- Range: 30-900 seconds (15 minutes)
4. Scaling Policies:
1.Target Value for Auto Scaling: The desired average metric value (such as CPU utilization or request count) for the Auto Scaling target. The Auto Scaling group will adjust the number of instances to keep this metric close to the target value.
Typical values: 50-80% CPU utilization, or a specific request count per instance depending on the application.