It is important to note that backups and data replication are not mutually exclusive. Both can play a critical role in disaster recovery planning, depending on your needs. RTO and RPO are part of the non-functional requirements (NFR) of a solution. NFRs capture conditions that are not directly related to the behavior or functionality of the solution, but describe the qualities that the systems must have or the environmental conditions in which the solution must remain effective. One trap I`ve seen many times is not involving the project budget owner in these discussions. Each project has a budget constraint. Period. When discussing RPO and RTO requirements, you should keep in mind that there is always a financial trade-off. What financial costs can you use to justify risk mitigation? Every company wants a near-zero RPO and RTO solution with five nines availability. However, this can be extremely costly. If the company can tolerate 4 hours of downtime, it doesn`t make sense to spend too much money on a solution with near-zero downtime.
Good design is also cost-effective and responsible design! Availability SLAs are often available with different tiers such as Gold, Silver and Bronze or Tier I, II and III. This allows you to differentiate yourself in your solution and take different measures. Especially within an SDDC, it is quite possible to provide different SLAs for workloads running in that SDDC. For example, temporary test systems typically don`t require PFTT and SFTT redundancy, so why replicate data multiple times across your vSAN cluster? Systems with built-in data replication, such as databases or MS Exchange, may not require storage-level redundancy. Modern applications that run stateless in sites behind a GLSB might not require a failover solution for site-level site recovery. These are all factors you need to consider in your design when taking steps to meet availability and recoverability requirements. These design qualities are almost always non-functional requirements because they specify HOW the system should behave (how fast, how safe it is, how recoverable it is, etc.). Functional requirements are about what a system needs to do (provide self-service capabilities to the enterprise, deploy a consistent hybrid cloud infrastructure, provide a single interface for management, etc.) Each design contains (or should include) J) RPO and RTO requirements.
Recovery point objective (RPO) determines your tolerance for data loss, typically in hours. Recovery time objective (RTO) describes how long you expect a failed system to be fully operational again. In a visual representation, it looks like this: If you look at the VCDX plan, it becomes clear that a VCDX design must contain non-functional requirements or design qualities, such as: So let`s take a look at the following diagram that illustrates the use of RTO and RPO: Unitrends provides the perfect technology stack to shape RPO and RTO in such a way that it perfectly matches your business needs. Source: blogs.vmware.com/virtualblocks/2018/09/30/more-srm-and-vsphere-replication-faqs/ Recovery Time Objectives (RTOs) indicate the time elapsed between the occurrence of an outage event and when affected resources must be fully operational and ready to support the organization`s objectives. Figure 1 shows the RTO metric. In the event of a disruption, no business can suffer data and reputation loss without RPO and RTO. You will have no idea how much and for how long your business can afford data loss. An effective RPO and RTO add a more pragmatic layer to your BCDR plan and make policies more effective.
Calculation variables: Based on the smallest number of variables, RPOs may be easier to calculate due to consistency in data usage. RTOs are a bit more complicated because recovery times depend on several factors, including analog delays and the day the event occurs. A shorter RPO means less data lost, but it requires more backups, more storage capacity, and more compute and network resources to run backups.