Azure Outages Reinforce Importance of a Disaster Recovery Plan

Sometimes you don’t have to lose your data outright to suffer major disruption to your business. With more companies outsourcing their IT infrastructure to IaaS providers like AWS and Azure, cloud reliability is becoming more of a concern.

In the past week alone, Microsoft Azure has suffered from two major outages: one on 9/9 bringing down cloud services for many customers in its North Europe, West Europe and India regions completely. Then on 9/15, there was a global DNS outage which again, left some Azure customers without access to their SQL databases, Virtual Machines and backups.

It’s not just Microsoft who is having these issues. Google Apps for Work users also went down for over an hour on 9/14, leaving US and UK users who rely on the cloud productivity platform in the lurch. In the past, AWS has also had hiccups up to and including complete outages, bringing down popular websites including Netflix, Reddit and IMDB.

What does this mean for you as an IT professional? First, you need to realize that no cloud is perfectly reliable and downtime is not a matter if, but when. Even a couple of hours of disruption can rack up millions in lost sales and/or productivity – according to an IDC report, the average hourly cost of an infrastructure failure is $100K/hour, or in the case of a critical application, $500K - $1MM/hour.

While some public clouds have a better reputation for uptime than others, the only sure-fire way to maintain business continuity is creating a disaster recovery plan. A disaster recovery plan, may include backup, replication, and failover to a different hot/warm site. Although services like Azure have their own backup facilities, your best defense is to keep backups on a different cloud entirely. Even when an IaaS provider has different availability zones, consider that outages like the recent ones with Azure frequently span across multiple regions.

The worst thing about a cloud outage is how you’re entirely at the mercy of the ops team at Microsoft, Amazon or Google to fix it. They have thousands of customers to serve, and unless you’re a major customer with a SLA, good luck getting in touch with anyone besides front-line support. With a private cloud, you could put “boots on the ground” to repair the faulty part of your infrastructure, but with a public cloud, you don’t have that access because everything beyond your virtual server is very opaque.

Microsoft attributed its DNS outage last Thursday to high traffic, illustrating how a neighbor overwhelmed by a burst of traffic in the public cloud (whether legitimate or a DDoS attack) can bring you down along with them.

Many companies have enjoyed cost savings and productivity gains from moving to the cloud, but it’s imperative to mitigate the risks as well. Using a Disaster-Recovery-as-a-Service (DRaaS) solution such as Asigra Cloud Backup can help you easily meet your recovery time objectives (RTOs). Key benefits include:

  • End-to-end solution protects traditional and cloud data sources
  • Seamlessly failover to an alternate datacenter with Virtual Disaster Recovery (VDR)
  • Flexible and scalable solution for companies of all sizes to meet business needs today and future

For more information, contact a Recoverability Specialist who will be happy to discuss your SaaS, IaaS and cloud-to-cloud backup needs.

Topics Discussed

Related Posts

Technology Defined: Disaster Recovery as a Service (DRaaS) vs Disaster Recovery (DR) Losing business critical data can cost companies millions of dollars and force some companies to go... 2 min read
Part 2: How Multiple Data Protection Systems Corrupt Recovery Processes Data is the crown jewels of every organization. Failure to protect it from loss, outages, theft,... 4 min read
Highstreet IT Case study Highstreet IT Solutions, an IT outsourcing company was looking to expand into managed services with... 1 min read