We should monitor so that we don't send out a faulty product.
The aspects we should monitor are:
- error logs
- budgeting
- uptime - access time - response time
- security breaches
- system test/health
- instance's health
- CPU utilisation %
- Memory utilisation %
4 Golden Signals
- Lantency - Request Service Time
- Traffic - User Demand
- Errors - Rate of failed requests
- Saturation - Overall capacity of the system
Three monitoring services in AWS
- Cloudwatch - Monitor AWS service
- SNS - Simple notification
- SQS - Simple queue service
- Go on instances page, find the instances you want to set the alarm on, and find the alarm column.
- Once found, click the plus sign and you should then fill out necessary information such as the percentage.
- After that is done, you will need to enter and confirm your email.
- Now you just wait for however long you set your period length and wait for an alarm once it has the hit the requirements.