Scaling policies (target tracking, step, simple) in AWS - Step-by-Step Execution

Process Flow - Scaling policies (target tracking, step, simple)

Start

↓

Monitor Metric

↓

Compare Metric to Target

↓

Target Tracking Policy

↓

Adjust Capacity to Match Target

↓

Update Resources

↓

Step Scaling Policy

↓

Check Metric Thresholds

↓

Scale Up/Down by Steps

↓

Update Resources

↓

Simple Scaling Policy

↓

Metric Breaches Threshold?

↓

Add or Remove Fixed Capacity

↓

Update Resources

↓

End

The system monitors a metric and applies one of three scaling policies: target tracking adjusts capacity to keep the metric near a target; step scaling changes capacity in steps based on thresholds; simple scaling adds or removes fixed capacity when thresholds are crossed.

Execution Sample

AWS

Metric = CPUUtilization
Target = 50%
If Metric > Target + 5%: Scale Out
If Metric < Target - 5%: Scale In

This example shows a target tracking policy that scales out if CPU usage is above 55% and scales in if below 45%.

Process Table

Step	Metric Value	Policy Type	Condition Checked	Action Taken	New Capacity
1	40%	Target Tracking	40% < 50% - 5%	Scale In	Capacity - 1
2	48%	Target Tracking	48% within target range	No Change	No Change
3	56%	Target Tracking	56% > 50% + 5%	Scale Out	Capacity + 1
4	70%	Step Scaling	70% > High Threshold (60%)	Scale Out by 2	Capacity + 2
5	30%	Step Scaling	30% < Low Threshold (40%)	Scale In by 1	Capacity - 1
6	55%	Simple Scaling	55% > Threshold (50%)	Scale Out by 1	Capacity + 1
7	45%	Simple Scaling	45% < Threshold (50%)	Scale In by 1	Capacity - 1
8	50%	All Policies	Metric at Target	No Change	No Change

💡 Metric stabilizes near target; no further scaling needed.

Status Tracker

Variable	Start	After Step 1	After Step 2	After Step 3	After Step 4	After Step 5	After Step 6	After Step 7	After Step 8
Metric Value	N/A	40%	48%	56%	70%	30%	55%	45%	50%
Capacity	10	9	9	10	12	11	12	11	11

Key Moments - 3 Insights

Why does the capacity not change at step 2 even though the metric is close to the target?

At step 2, the metric is within the target range (48% is between 45% and 55%), so the target tracking policy decides no scaling action is needed, as shown in execution_table row 2.

How does step scaling decide how much to scale?

Step scaling uses defined thresholds to decide scaling amount. For example, at step 4, metric 70% exceeds the high threshold 60%, so it scales out by 2 units, as seen in execution_table row 4.

What is the difference between simple scaling and target tracking in scaling decisions?

Simple scaling adds or removes a fixed amount when a threshold is crossed, without aiming to maintain a target metric. Target tracking adjusts capacity continuously to keep the metric near a target, as shown by comparing steps 1-3 (target tracking) and steps 6-7 (simple scaling).

Visual Quiz - 3 Questions

Look at the execution table, what is the capacity after step 4?

A 10

B 12

C 11

D 9

💡 Hint

Check the 'New Capacity' column at step 4 in the execution_table.

At which step does the metric first cause a scale in action?

A Step 1

B Step 3

C Step 5

D Step 7

💡 Hint

Look for 'Scale In' in the 'Action Taken' column in the execution_table.

If the metric at step 6 was 45% instead of 55%, what would the simple scaling action be?

A Scale Out by 1

B No Change

C Scale In by 1

D Scale Out by 2

💡 Hint

Refer to the simple scaling logic in execution_table rows 6 and 7.

Concept Snapshot

Scaling policies adjust resource capacity based on metrics.
Target tracking keeps metrics near a target by adjusting capacity dynamically.
Step scaling changes capacity in steps when thresholds are crossed.
Simple scaling adds or removes fixed capacity on threshold breaches.
Each policy monitors metrics and updates capacity accordingly.

Full Transcript

Scaling policies help cloud resources adjust automatically to demand. Target tracking policies watch a metric like CPU usage and add or remove resources to keep it near a set target. Step scaling policies check if the metric crosses certain thresholds and then scale up or down by defined steps. Simple scaling policies add or remove a fixed number of resources when a threshold is crossed. This visual trace shows metric values, policy decisions, and how capacity changes step by step, helping beginners understand how each policy works in practice.