Alarm actions (SNS, Auto Scaling) in AWS - Step-by-Step Execution

Process Flow - Alarm actions (SNS, Auto Scaling)

Metric monitored

↓

Alarm threshold reached?

No→Wait

Yes↓

Trigger Alarm

↓

Send SNS notification

→Subscribers notified

↓

Trigger Auto Scaling action

→Scale up/down instances

The system monitors a metric, triggers an alarm when a threshold is crossed, then sends notifications and/or scales resources automatically.

Execution Sample

AWS

Alarm:
  Metric: CPUUtilization
  Threshold: > 70%
  Actions:
    - SNS Topic
    - Auto Scaling Policy

This alarm watches CPU usage, and when it goes above 70%, it sends a message and adjusts server count.

Process Table

Step	Metric Value	Condition (>70%)	Alarm State	Action Taken
1	45%	False	OK	No action
2	65%	False	OK	No action
3	72%	True	ALARM	Send SNS notification, Trigger Auto Scaling
4	75%	True	ALARM	Send SNS notification, Trigger Auto Scaling
5	68%	False	OK	No action

💡 Metric drops below threshold at step 5, alarm state returns to OK, actions stop.

Status Tracker

Variable	Start	After 1	After 2	After 3	After 4	After 5
Metric Value	N/A	45%	65%	72%	75%	68%
Alarm State	OK	OK	OK	ALARM	ALARM	OK
Action Taken	None	None	None	SNS + Auto Scaling	SNS + Auto Scaling	None

Key Moments - 2 Insights

Why does the alarm only trigger actions when the metric is above 70%?

Because the condition (>70%) is only true at steps 3 and 4, as shown in the execution_table rows where Alarm State changes to ALARM and actions occur.

What happens when the metric falls back below the threshold?

At step 5, the metric is 68%, condition is false, so the alarm state returns to OK and no actions are taken, as seen in the execution_table.

Visual Quiz - 3 Questions

Look at the execution_table, what is the Alarm State at step 2?

A OK

B ALARM

C INSUFFICIENT_DATA

D UNKNOWN

💡 Hint

Check the 'Alarm State' column at step 2 in the execution_table.

At which step does the alarm first trigger Auto Scaling actions?

A Step 1

B Step 3

C Step 2

D Step 5

💡 Hint

Look at the 'Action Taken' column in the execution_table to find when Auto Scaling starts.

If the threshold was changed to > 60%, how would the alarm state at step 2 change?

A It would become INSUFFICIENT_DATA

B It would remain OK

C It would change to ALARM

D It would be UNKNOWN

💡 Hint

Step 2 metric is 65%, which is above 60%, so condition would be true.

Concept Snapshot

Alarm actions watch a metric and trigger when thresholds are crossed.
Actions include sending SNS notifications and Auto Scaling adjustments.
Alarm state changes from OK to ALARM when triggered.
Actions stop when metric returns below threshold.
Use alarms to automate responses to system changes.

Full Transcript

This visual execution shows how an alarm monitors a metric like CPU usage. When the metric goes above 70%, the alarm state changes to ALARM. This triggers two actions: sending a notification via SNS and triggering an Auto Scaling policy to adjust server count. When the metric falls back below 70%, the alarm state returns to OK and actions stop. The execution table tracks metric values, alarm state, and actions step-by-step. This helps beginners see exactly when and why actions happen in response to metric changes.