Businesses today are all about “agile.” For storage professionals, this means delivering ultra-low latency alongside high throughput and high input/output (I/O) operations per second (IOPS) to accelerate mission-critical application performance as much as possible. All-flash storage arrays have solidified their place in production data centers by meeting this need. What is less frequently recognized is that users expect the same fast levels of application performance when in a failover, disaster recovery state.
Naturally, the business cannot slow down in the event that a disaster recovery state is declared. At the same time, the precedent has been set that aggressive levels of performance are not only possible, but increasingly that they are table stakes, as flash storage becomes much more common in production data centers. The business expects flash-levels of performance even during a period of recovery. To meet these expectations, it now makes sense to implement a manageable, strategically sized tier of flash storage in the secondary data center.
Declining price points of SSDs – to the point where they are in fact no longer vastly more expensive than HDDs – makes it feasible today to deploy flash capacity in the secondary data center. This capacity may be deployed in the form of an all-flash system, or it may be deployed in the form of a hybrid system, depending on the enterprise’s performance service level agreements (SLAs) during recovery. From this vantage point, flash is emerging as an ideal complement from a storage media perspective to the instant recovery-like features that are becoming more common in disaster recovery software.
Naturally, SSDs still cost more than HDDs, so copy data must be effectively managed in order to control the flash capacities – and as a result costs – in the backup environment. Working in IT’s favor is the fact that the vast majority of recoveries (Storage Switzerland sees upwards of 95%) come from the most recent backup. This means that the lion’s share of copy data can be tiered to a less expensive storage media, such as an on-premises or public cloud-hosted object store. To truly optimize flash capacity utilization, this tiering must happen as quickly as possible.
To learn more, access Storage Switzerland’s on demand webinar with Veeam and NetApp “Will Your Backup Architecture Meet Tomorrow’s SLAs? 3 Steps to Make Sure!”. All registrants receive a copy of our latest White Paper “How to Design a Modern Data Protection Architecture.”