Having an effective disaster recovery plan and ensuring business continuity involves many facets of planning and consideration. It also requires putting the correct mechanisms in play that allow effectively meeting the goals as outlined in the business continuity/disaster recovery (BCDR) plan.
Having the right data protection solution that allows meeting the BCDR objectives effectively and efficiently is key for organizations today who are pressed to meet the high-pressure demands of the demanding “always on” data requirements.
Today’s businesses are no longer housed in on-premises datacenters with simple network topologies. Businesses today are embracing the public cloud to make use of cutting-edge technology and advanced offerings by today’s public cloud vendors. Organizations today are now very hybrid in nature. All of these facets weigh into designing proper backup and recovery mechanisms.
Following the 3-2-1 best practice backup methodology requires that organizations today diversify where backup data is stored. This requires that backups are not stored along with production infrastructure. Additionally, businesses must provide the means to bring up mission-critical VMs as quickly as possible. To accomplish both of these objectives, there are a couple of backup technologies that need to be implemented – offsite DR and VM replication.
Let’s take a look at
- What is the difference between backup and replication, and
- Why both are needed
Why Are Offsite Backup Copies Important?
When thinking about effective business continuity and disaster recovery plan, organizations must think about the worst-case scenario to be fully prepared. This means that businesses need to think not just about losing single or multiple VMs, but instead, at the scale of losing an entire site and so forth. The 3-2-1 backup rule provides excellent recommendations as to how backups of data need to be arranged and diversified. The 3-2-1 backup rule states that you need at least (3) copies of your data on (2) different kinds of media with at least (1) backup stored offsite.
Why the diversity and multiples of backups?
When thinking about the worst-case scenario for most organizations of losing an entire production site, it would be disastrous for data to have both the production data and the backup data stored in the same location. You can imagine the catastrophe that would unfold if both types of data were destroyed by a disaster. This underscores the importance of having backup data or at least a copy of backup data stored elsewhere besides the main production site or location. This ensures that even a large-scale disaster that may destroy the primary production location, would most likely not also destroy an offsite location, especially if these are stored in relatively diversified geographic locations.
As mentioned in the outset, many businesses today are making use of the public cloud and public cloud storage for many different business use cases. A great fit for public cloud storage is for offsite DR purposes since public cloud storage can infinitely scale and organizations can have relatively large amounts of storage that is very cost-effective. Public cloud storage is often a great fit from a data diversity perspective since organizations can make use of public cloud storage that is easily in a different region than production data. This allows businesses to not have to invest in the tremendous capital expense of building facilities and infrastructure to house a DR facility with private storage.
Public cloud storage allows businesses to transition these expenses into much more palatable operational expenses. This often provides a great way for startups who often do not have the initial funding to build out the network and other infrastructure that is needed to have a proper DR facility for storing backup data. Organizations wishing to further diversify and augment existing DR and offsite private facilities can also utilize public cloud storage to provide an additional location to copy backup data and provide further resiliency.
Upon reviewing the 3-2-1 backup best practices strategy, the offsite copy of backup data helps to satisfy at least a portion of all of the requirements. Offsite backups provide at least one of the (3) copies of data. Tape storage for offsite backup data or public cloud storage can easily qualify as one of the (2) different media types. By definition, offsite copies of backups satisfy the (1) copy of backup data offsite requirement.
When thinking about experiencing an entire site failure, businesses want to be able to recover production data as soon as possible depending on their RPO and RTO SLAs. When thinking about site failure and recovering access to production data as quickly as possible, replicated virtual machines sitting in a DR or secondary location is often the perfect means to quickly recover from a site-level failure.
What is virtual machine replication and how does it play into the overall business continuity and disaster recovery strategy?
Virtual Machine Replication – Key to Overall Site-Level Protection
Backups of virtual machines are key to providing the typical recovery needed for most disaster recovery situations where either a single virtual machine must be restored or files within a single virtual machine are in need of restoration. However, once the scope of a disaster transcends beyond several virtual machines that have become unavailable to any number of reasons, recovering these VMs by a simple restore process most likely will not meet RTO expectations. Additionally, if a disaster is large enough in scope where an entire site is unavailable due to an extensive disaster, recovery by restoring data may simply not be possible.
Virtual machine replication provides site-level protection for these types of situations. When replication is enabled on a virtual machine, an exact, identical copy of the virtual machine is created in the DR or secondary replication target. Then, with each consecutive replication interval, the incremental changes are synchronized with the replication target virtual machine. Depending on the RPO of site-level SLAs, replication intervals can be adjusted accordingly.
In the event of a site-level failure, the production virtual machines are “failed over” to the replicated copies of the virtual machines in a DR or secondary location. DNS and other network settings are quickly changed for the re-provisioned, powered-on virtual machines, allowing connectivity to the newly powered on copies of the VMs. This allows quickly recovering most of the data and the connectivity to the virtual machines extremely quick compared to the prospect of having to restore all the production data first, and then power on the restored copies of the production virtual machines in a DR facility.
While virtual machine replication is a great mechanism to allow quickly recovering access to production virtual machines and business-critical data, there are downsides to consider with replication. Replication requires an entire set of infrastructure that is housed at an offsite location that is comparable to the production infrastructure or is sized to run very similarly to production infrastructure from a performance standpoint. This means from a Cap-Ex standpoint, organizations are having to purchase basically twice the infrastructure as a target for replicated virtual machines for production workloads being replicated across. From an expenditure perspective, having a DR facility that basically is not utilized until a disaster strikes is a hard financial cost to justify for many organizations.
Again, public cloud infrastructure is allowing businesses to offset the tremendous expenditures as relates to physical infrastructure and facilities required to provision a proper DR facility. By making use of the public cloud provider’s Infrastructure-as-a-Service or IaaS offerings, businesses are able to shift these expenses into a much more efficient operational expense model.
Whichever option that businesses decide to choose, replication is an extremely important part of the overall business continuity/disaster recovery strategy for protecting production data and workloads. It fits into the 3-2-1 backup strategy by providing an additional copy of backup data in the form of a pre-restored virtual machine that is replicated and synchronized at set intervals to provide relatively up-to-date copies of business-critical data. Additionally, replication provides a means to satisfy the offsite backup requirement of the 3-2-1 backup best practice methodology.
Final Thoughts
Organizations today must protect production data as this is key to providing resiliency to both data and services. Two key mechanisms to provide this resiliency for business continuity/disaster recovery purposes are the offsite copy of backup and replication of virtual machines. Solutions such as Vembu BDR Suite provide extremely effective data protection mechanisms so that organizations today can meet the demands of the 3-2-1 backup best practice methodology. Vembu provides the Offsite DR server in both an on-premises (self-hosted offsite) version and a Vembu CloudDR solution which is a fully-hosted cloud solution for providing a target for offsite copies of backup data. By utilizing both backup and replication, businesses today can have a powerful means to ensure multiple copies of backup data as well as site-level resiliency in the event of a site-level failure.
Follow our Twitter and Facebook feeds for new releases, updates, insightful posts and more.
