Availability

A system is available if it is able to perform its functionality (uptime/total time). 6 Reliability 2023112801224646 Reliability implies availability but availability does not imply reliability.

Percentage of availability = (total elapsed time – sum of downtime)/total elapsed time SLA - 99.999

Availability (%)	Downtime (per year)
90% (one nine of availability)	36.53 days
99% (two nines of availability)	3.65 days
99.9% (three nines of availability)	8.77 hours
99.95% (three and a half nines of availability)	4.38 hours
99.99% (four nines of availability)	52.60 minutes
99.995% (four and a half nines of availability)	26.30 minutes
99.999% (five nines of availability)	5.26 minutes

💡 To increase availability, you need redundancy.

Challenges

• replication
• customer redirection DNS, load balancer
• types of high availability
  • active-passive
  • active-active
    • 💡 Stateless applications work better for active-active systems

References

1. https://www.bmc.com/blogs/reliability-vs-availability/
2. AWS EssentialsModule 6
3. AWS whitepaper: High availability and scalability on AWS(opens in a new tab)
4. AWS documentation: Reliability Pillar – AWS Well-Architected Framework(opens in a new tab)
5. AWS website: Amazon EC2 Auto Scaling
6. https://www.tryexponent.com/courses/system-design-interview/fundamentals-system-design/system-design-principles