Reduce Recovery Point Objectives & Recovery Time Objectives (RPO & RTO)
There is nothing worse than having to re-enter data or lose ideas and intellectual property because of an unplanned server or site outage. Nobody wants to lose data. Recovery Point Objectives refer to how much data is actually being lost. Recovery Time Objective refers to how long it takes to get the data back online. Disaster Recovery solutions help customers by providing the least data lost and the lowest recovery times. The faster the services can resume and the data is available, the better it is for the organization.
A couple of common solutions that are used to protect data are hardware- and storage-based replication. Hardware- and storage-based replication solutions fall short when protecting the data because they leverage storage snapshots. Storage snapshots introduce unwanted overhead to the production systems, plus create a big network penalty when replicating the snapshots to the remote site and worst of all, they create a wide gap between data protection points.
When using a snapshot-based replication solution, the process includes:
- Waiting for the scheduled replication time
- Waiting for the built-up queue of data changes to replicate
- Finally, the data is written to storage at the target site
It’s no surprise that production RPOs for expensive snapshot-based solutions hope to achieve an RPO of about an hour. High RPO means more lost data.
How Zerto became the standard for data protection and failover orchestration
Zerto Virtual Replication doesn’t use snapshots because they are clearly too inefficient. Instead, Zerto was designed to be a highly efficient and easy-to-use solution, yet deliver the most aggressive RPOs possible.
ZVR is software-based and uses asynchronous replication at the hypervisor level. Software-based asynchronous replication meets the efficient design goals in many ways. It is much cheaper, more practical and easier to deploy and support than cumbersome hardware-based synchronous replication solutions. Yet ZVR achieves nearly identical replication benefits found in expensive hardware-based full synchronous replication solutions.
Zerto cleverly uses the asynchronous replication framework of Continuous Data Protection (CDP) at the hypervisor level, not the storage level, to capture the data before it is even written to the storage layer in order to be the closest to zero RPO as possible. ZVR replicates data as it passes through the hypervisor nearly simultaneously as it is being written to the primary site’s storage. So it provides the most aggressive RPOs in the market today. It is not unusual to have a 10 second RPO with Zerto Virtual Replication in real production deployments. Low Recovery Point Objectives equals less lost data.
Zerto Enables the Fastest Recovery Time Objectives (RTOs) in the Industry
When the servers are down, everyone knows it. Server downtime is the most obvious event in an organization. The business impact is substantial as teams of people are unable to perform their work. This often results in longer-lasting impact than the downtime event itself. Customer confidence, lost opportunities and employee morale are all negatively impacted.
Studies repeatedly show that server downtime isn’t usually caused by natural disasters, but actually human caused outages. Being able to recover a specific group of servers that support an application is just as important as being able to quickly recover an entire site.
How Zerto’s software drives down RTOs
There are three primary ways to reduce the recovery time of servers:
- Have the servers in affinity groupings that logically go together, either by tiers or by service or application support. Zerto calls these Virtual Protection Groups (VPGs).
- Integrating the orchestration and replication tools for administrative efficiency
- Thoroughly and routinely perform failover tests. Perfect the boot order, re-ip addressing, and scripts well ahead of when an outage occurs. Have the failover push-button ready.
Virtual Protection Groups:
ZVR has a powerful feature called Virtual Protection Groups (VPGs). VPGs enable the grouping of VMs that make an application or service function. The servers within the VPG are protected together, with write-order fidelity between the VMs in the VPG. The write-order fidelity is important because Zerto has a point-in-time (PIT) journal that spans as far back as 14 days. The recovery points are very granular and there is a point entered into the journal every few seconds. The benefit of the write-order fidelity comes into play when a VPG is failed over to the recovery site. All of the servers are recovered together at the exact point in time because of the VPG write-order fidelity feature of ZVR. Having the servers logically grouped and failed over together drives down the Recovery Time Objectives.
Combined Orchestration and Replication Tool:
There is an elegance to having both the orchestration and the replication tool combined that isn’t found on many other solutions that are essentially cobbled together components. For example, when configuring the failover, ZVR configures the reverse protection scheme, meaning it will automatically replicate back to the source site and have the failback configurations defined based on the location of the production VMs. ZVR also automates the end-user acceptance confirmation based on their SLA. The user interface provides both replication and virtual machine protection information in the same dashboard.
It is impossible to perfect the disaster recovery process or determine the duration of a downtime without regular DR testing. Historically, disaster recovery testing either didn’t occur or happened so infrequently that the test results and organizational preparedness were effectively worthless. Zerto features non-disruptive testing of the failover, giving organizations insight to and management of all of the moving parts of a failover operation such as re-IP addressing, pre- or post- recovery scripting and server boot order – all done with no disruption or impact to the production environment.
Non-disruptive testing is the key to lowering Recovery Time Objectives for both disaster recovery operations as well as migrations. There is no dependence on storage synchronization between sites because ZVR uses software-based CDP replication. The data is already at the recovery site, so there is zero time lost on waiting for storage to do a final delta synchronization. The migration can be fully tested prior to the downtime window and the migration can be done all at once or systematically in a series of application migrations based on Zerto’s virtual protection groups.
To have a cost-effective, yet lowest RPOs possible solution, the mechanism that replicates the data has to be as close as possible to the instant that it is being written from the virtual machine, through the hypervisor to the storage. That is exactly where ZVR replicates data as the point of origin. Being software-only removes all of the physical barriers to deployment and makes installation and supportability easier than any other solution on the market. Ease of use and support means less data loss.
Driving down RTOs means less downtime for the end users and customers. For the lowest possible recovery times, the tool being used needs to allow for repeated, non-disruptive testing and be powerful enough to push RTOs to the lowest level possible. Zerto Virtual Replication delivers all of these features and enables systematic DR testing, DR, and migrations.