Damaging storms in Sydney, Australia this past weekend took out roads, buildings, and power for more than 226,000 homes and businesses. The storms also shut down Amazon Web Services’ Sydney availability zone for hours, causing disruption for many leading Australia-based businesses.
Domain, Domino’s, and The Iconic all experienced extended downtime when EC2 and EBS instances went offline during the storms. According to the Australian Financial Review, “It is understood that power was lost in a number of systems in its data centre, before being restored about 90 minutes later. It then took hours for some systems to be rebooted and brought back online.”
The outages experienced by the leading public cloud provider serve as an important reminder: The cloud is not immune to failures. This is hardly a warning to avoid the cloud. On the contrary, the benefits cloud delivers can be immense. But just because your architecture has migrated away from on-premises does not mean you’ve eliminated the risk of damaging downtime. There are a couple strategies for mitigating risk with cloud architectures.
First, organizations can take a multi-zone or multi-region approach to infrastructure. For example, some companies in Australia were able to click over to the Singapore region when the Sydney zone went down. These business continued to operate without disruption.
The second strategy is to have another cloud provider ready for failover. Many companies have adopted this multi-cloud approach to infrastructure. It allows companies to leverage the best pricing and strengths of each vendor. But it also protects against cloud failure if one vendor suffers an outage. While multi-region redundancy may be easier to execute, both strategies make sense.
Smart developers build applications with failure in mind. You have to assume infrastructure will never be 100% reliable. It doesn’t matter if the infrastructure is on-prem or in the cloud. Events like the AWS outage in Sydney also emphasize the need for IT infrastructure monitoring solutions. But these solutions must provide visibility across multi-cloud and hybrid IT architectures.
ScienceLogic is the perfect example of a hybrid IT monitoring solution. It provides line of sight when infrastructure performance or availability suffers. It also provides insight about which cloud environments are best-suited for handling business-critical services.
It also automatically discovers and maps resource dependencies across on-prem and cloud infrastructure. You can view all your cloud infrastructure in one screen, across all regions and zones. Then drill into specific areas of concern when troubleshooting issues that impact service delivery. (Learn more about ScienceLogic’s cloud monitoring capabilities)
The Sydney storm won’t be the last event that threatens cloud architectures. But it serves as a wake-up call for organizations who have not yet diversified their infrastructure. Taking a multi-zone, multi-region, or multi-cloud approach to supporting business-critical applications can help avoid costly downtime.