AWS Challenges and Issues

AWS Experiences Another Outage in US-EAST-1 Region

Amazon Web Services (AWS) encountered its third service disruption this month. A power failure within a data center in the US-EAST-1 region initiated the problems today.

The initial reports surfaced around 7:30 a.m. ET, and the repercussions are still being felt as of 1 p.m. ET. AWS is currently addressing difficulties with several services within this region, notably the EC2 compute service and associated networking capabilities.

Details of the Disruption

According to an 8 a.m. ET update, the company confirmed a power loss impacting a single data center within Availability Zone (USE1-AZ4) in the US-EAST-1 Region. This has resulted in impaired availability and connectivity for EC2 instances located within that specific data center and Availability Zone.

Elevated error rates were also observed when launching new instances within the affected zone. However, connectivity and power to other data centers within the same Availability Zone, or to other Availability Zones within the US-EAST-1 Region, remain unaffected.

AWS recommends failing over from the impacted Availability Zone (USE1-AZ4) for those who have the capacity to do so.

A Pattern of Recent Outages

While isolated incidents are expected in complex cloud environments, AWS has been experiencing outages on a weekly basis. A similar disruption in the US-EAST-1 region occurred on December 7th, stemming from a networking issue.

Furthermore, on December 17th, an outage affecting connectivity between two West Coast regions caused service interruptions for companies like Netflix, Slack, and Amazon’s Ring.

These events are particularly noteworthy given AWS’s recent emphasis on the resilience of its cloud infrastructure at the re:Invent conference.

Mitigation and Future Considerations

Users can mitigate the impact of such outages by designing systems that automatically fail over to a geographically distinct region. However, this approach can significantly increase costs.

Ultimately, the responsibility for providing a stable platform rests with AWS. Whether these outages are due to unfortunate circumstances or underlying systemic issues remains to be seen.

For services currently hosted in the US-EAST-1 region, exploring alternative locations may be a prudent consideration.