8-Hour Azure Outage Highlights Need for Independent Disaster Recovery Solution for Cloud Workloads

By John Maxwell

Overview

Infrastructure as a Service (IaaS) has become part of the mainstream compute infrastructure for both small and large businesses. The benefits are many from no CAPEX to “infrastructure on-demand” to support growing or seasonal workloads. But when public cloud services fail, businesses can find themselves in a “disaster recovery” scenario just as they would if they had a disruption in their own datacenter.

Such is the 8-hour outage that occurred with Microsoft Azure, where a global outage impacted Azure Virtual Machines and related services.

Azure Disaster Recovery

Outages such as the recent Azure incident are rare, but they can happen. This is why Managed Service Providers (MSPs) responsible for their clients' workloads in IaaS environments need to have a Business Continuity and Disaster Recovery (BCDR) plan in place, just as they would for their customers whose workloads run on-premises or in a private cloud (such as an MSP’s co-location facility).

While many of the “best practices” for BCDR are the same for on-premises, there are several nuances for IaaS workloads.

  1. Adhere to the 3-2-1 methodology where you have multiple backup copies that are offsite.
    1. While solutions exist to perform backup within cloud services such as Azure, this exposes clients to undue risk when there is a broad and sustained outage.
  2. Guarantee that the backups of the primary cloud services provider are recoverable within an optimal timeframe.
    1. Microsoft does not take sole responsibility for protecting data in their cloud server. Instead, that responsibility is shared between Microsoft and the customer (SMBs and their MSP).
  3. Provide the same level of computing resources needed to run the production workloads for clients that were running in the affected environment.
    1. If Azure is down, how and where can you recover?
  4. Provide the highest level of recovery speed (aka RTO) and granularity to recover from (RPO).
    1. Where do you find spare capacity to recover one or more production virtual machines and how current are the available backups?

Datto’s answer to Azure Recovery

Datto provides a purpose-built solution for MSPs to provide all-in-one BCDR for their clients who run their infrastructure in Microsoft Azure. The solution, Datto Backup for Microsoft Azure, is unique in that while it runs as a cloud service to seamlessly protect Azure-based workloads, it utilises the immutable Datto Cloud for backup retention and disaster recovery. Using the best practices listed above, let’s see how Datto Backup for Microsoft Azure ranks

  1. While Datto Backup for Microsoft Azure is a service within Azure, all backup snapshots are replicated to the immutable Datto Cloud with $0 in egress charges
    1. By storing backup copies outside of Azure, MSPs can recover their client's data outside of Azure in case of a broad outage.
  2. MSPs can perform instant virtualisation of their clients' systems in the Datto Cloud in seconds and get clients up and running in minutes - not days.
    1. Datto Backup for Microsoft Azure provides hourly backups of Azure workloads that are stored in the Datto Cloud providing the ability to restore from a backup that is an hour old or less.
  3. Datto Backup for Microsoft Azure is built with multiple layers of safety to ensure backups can be recovered.
    1. All backups are application-consistent, meaning applications running in the affected VMs like SQL Server will be able to recover with referential integrity.
    2. The backup file snapshots themselves cannot be infected with ransomware.
    3. Backup file snapshots are replicated to the secure Datto Cloud via AES 256 encryption.
    4. Backups are encrypted at rest in the Datto Cloud and all data in the Datto Cloud is monitored 24/7 by information security professionals.
    5. Datto utilises a multi-step backup verification process to ensure a backup is recoverable.
      1. This includes patented Screenshot Verification to ensure the system is bootable and optional advanced screenshot verification to ensure an application can come up and be accessed
  4. The exabyte-class Datto Cloud is available 24/7/365 with dedicated customer support to help MSPs recover their client systems.
    1. The scalable Datto Cloud is built to provide MSPs with the compute power they need to recover one or all of their client’s systems.

On top of the best practices outlined above, one of the key tenets of a best-in-class BCDR solution is the ease of use. When a technician is recovering one or more systems, mistakes can occur which is why the intuitive Datto Backup for Microsoft Azure recovery launchpad walks you through the recovery process step by step.

8-Hour Azure Outage Highlights Need for Independent Disaster Recovery Solution for Cloud Workloads



Conclusion

The growth of workloads in public clouds will not soon abate nor will the attractiveness of IaaS as a means of running workloads for a growing number of SMB organisations. MSPs must be at the forefront of providing best-in-class BCDR for their clients no matter where their data resides. Datto Unified Continuity is the industry-leading solution for MSPs that spans on-premises and cloud to laptop and Office 365.

The MSP’s Guide to Business Continuity and Disaster Recovery on Azure

We created this guide to business continuity and disaster recovery (BCDR) on Azure to help MSPs prepare.

View the Resource

Suggested Next Reads