Dell EMC PowerStore has the storage industry buzzing with its data-centric architecture, integrated intelligence and machine learning, and adaptable infrastructure that provides flexible deployment options for both traditional and modern […]
Dell EMC PowerStore has the storage industry buzzing with its data-centric architecture, integrated intelligence and machine learning, and adaptable infrastructure that provides flexible deployment options for both traditional and modern workloads. Within this data-centric architecture, PowerStore prides itself on a no-compromise approach to data-services, offering benefits such as leading data reduction, without impacting performance. PowerStore leverages Intel’s Quick Assist Technology for accelerating compression without impacting the arrays maximum potential. In keeping with this “no compromise” focus for data services, PowerStore also delivers metro Active-Active capabilities with zero performance impact on the storage controllers or data path. Welcome PowerStore metro node! Metro configurations are the next step in a truly disaster resilient architecture, enabling business continuity – without intervention!
To understand the full benefits of PowerStore metro node we must first start by defining a few key industry terms, like RPO and RTO.
First there is RTO – Recovery Time Objective. RTO answers the question of “how long does it take to recover?” This is the time it takes to be fully up and running once the recovery process is initiated. If I have an RTO of two hours, then I must have data practices in place that allow me to have everything back up and running within a two-hour window once the recovery is started.
Second is RPO – Recovery Point Objective. RPO answers the question of “How much data will I lose after the recovery?” The RPO is the delta, or difference, between recoverable data sets. This might mean recovering from a backup, a snapshot, or a replicated data set at a remote location. Regardless of the recovery method, the RPO is the precise point in time you can recover to in the life of a data set. For example, if I am taking a volume level snapshot every 15 minutes, my RPO would be 15 minutes. I could potentially lose up to 15 minutes worth of data, the delta between snapshots.
RPO/RTO – but what about DTO?
Although RTO and RPO are very common terms in the storage industry, there is a third term I want to introduce. This term, credit to my friend Olly Shorey, is DTO: Decision Time Objective. DTO answers the question of “How much time is needed to make the decision to failover?” This is critical and often goes overlooked. How much time elapses from the time I recognize there is a problem and go through the necessary management / corporate approval process for a recovery to be initiated? This is DTO.
With the 3 key terms defined, we can take a closer look at the modes of replication commonly offered by storage systems today.
The first and most common mode of replication is asynchronous (asynch). Asynchronous replication is a periodic transmission of data between two different storage systems. This would mean that my RTO/RPO/DTO will always be greater than zero.
Second is synchronous (synch) replication which is less common in the industry. Synchronous replication means that both storage systems are fully mirrored. This would provide an RPO of zero BUT RTO/DTO would always be greater than zero because a failover and decision process would have to be manually initiated.
Third is metro, sometimes called active-active. In a metro replication configuration both storage systems are fully mirrored, and data is fully accessible at both locations simultaneously. This means we have an RPO/RTO/DTO that is always equal to zero. There is no point in time to recover to, there is no time associated with recovery, and there is no decision to make because the systems make the decision in an autonomous fashion.
Since Metro is the topic of conversation here, I want to provide a quick rundown of the advantages of PowerStore with metro node. PowerStore metro node offers true active-active locations with genuine metro capabilities – RPO/RTO/DTO = ZERO. Customers leveraging PowerStore with metro node have the ability to accomplish not only an RPO and RTO of zero, but also DTO of zero too! There is no decision that must be made because this system makes the decision for you, with no delay, based on industry leading technology. Customers can implement PowerStore metro node with the full confidence that data will always be available and accessible regardless of location (within metro geographical boundaries).
This intelligent decision-making process happens by having a witness, that is installed as a virtual machine. This witness prevents split-brain scenarios. Split-brain scenarios happen when the two arrays in a metro configuration lose communication with each other but could still be online. The witness prevents these situations and autonomously makes the decision to failover for you based on rules you define. This is critical when understanding DTO or Decision Time Objective. You never have to worry about manual intervention with PowerStore metro node and our autonomous witness technology.
As if metro node for PowerStore wasn’t news enough, it gets better. PowerStore is orderable and deployable with this advanced Metro capability today! Yes, PowerStore customers can recognize all the benefits of PowerStore’s scalability, date reduction, and performance in an easy to use no impact Metro configuration.
Not only is PowerStore metro node available today, there are several key technology benefits:
Metro node can support differing array types – you get to use the right technology at the right place at the right time. Customers can leverage Dell EMC technologies likeUnity XT with metro node as well.
Metro node also works on a granular workload selection. This allows you to select what you want to use metro capabilities for and what you do not. You are not forced to go all or nothing. This minimizes cost and negates the need to have identical system types or investments at each location.
Metro node has ZERO Performance impact on the array. You will never have to worry about array overhead or impact on production IO while trying to service active-active scenario. In many cases, it can increase the performance!
Metro node allows for true metro by allowing simultaneous writes at both sites.
Metro node has an Ansible module to fully support enterprise infrastructure automation.
PowerStore metro node is providing a service to customers that is unmatched in the industry. Other vendors metro solutions may not meet application timeout criteria or may add as much as 30% overhead to the storage array, therefore impacting overall performance. Take note of the “psuedo-metro” example in the above image. Some competing solutions call their solution metro when actually it is “psuedo-metro”. What do we mean by pseudo metro? Well, active/passive in effect. One site doing the work, and another site just standing by. The second site simply acts as a proxy for each workload, redirecting IO to the primary. With PowerStore metro node you will never have these challenges but will have the industry’s most robust active-active solution. PowerStore metro node provides the ability to write at both locations simultaneously – this is TRUE-metro and not “psuedo-metro”
In summary, PowerStore metro node leverages industry leading technology to seamlessly create TRUE active-active metro capabilities with zero performance impact, autonomous decision making, and enterprise infrastructure automation with Ansible. The result is maximum application uptime and availability without having to worry about the integrity of your data or the overall impact to your storage infrastructure.
A demo of PowerStore with metro node integration is available here
Leave a Reply Cancel reply