Purestorage vs Dell EMC PowerMax



Whatta a Fight!

Dell EMC PowerMax

The Dell EMC PowerMax family consists of two models, PowerMax 2000 and 8000, which support maximum capacities of 500TBe and 4PBe respectively, in half of a full rack for the PowerMax 2000 and up to two full racks for the PowerMax 8000. The PowerMax 8000 scales from one to eight dual-controller engines, now called Bricks, while the PowerMax 2000 supports either one or two Bricks with slightly slower and fewer CPU cores than the PowerMax 8000. The PowerMax evolved from EMC’s previous VMAX system, which supported a mix of disk-based and flash storage, and which in turn evolved from EMC’s Symmetrix and Symmetrix DMX systems; these originally supported high-latency HDD media. EMC’s first use of flash technology was with the addition of flash drives to DMX4 systems in 2008. While EMC’s original Symmetrix architecture was designed to overcome the latencies of magnetic, rotating HDDs, in contrast to Pure’s FlashArray, which is now in its fourth generation using NVMe, the Dell EMC PowerMax architecture has been retrofitted, and now also uses NVMe to communicate to its back-end storage devices – the first generation of its architecture to do so.

Purestorage X Family

The Pure FlashArray™ family consists of five different models: the //X10, //X20, //X50, //X70, and //X90 – scaling from 20TBe all the way to 3PBe in 6U using a highly available, dual-controller architecture with the ability to non-disruptively scale from one model to the next as performance and capacity requirements increase. From its inception, the Pure Storage architecture was purpose-built as an all-flash storage array, and is engineered to be able to non-disruptively upgrade with data in place and without the need for data migrations.

Although both architectures provide an all NVMe flash storage system that is both highly available and performant, there are many key differences to consider. This article will examine 5 essential criteria to consider when evaluating an enterprise-class all-flash storage array:
• Storage Efficiency
• Storage Management and Automation
• High Availability
• Environmentals
• Total Cost of Ownership (TCO)

Storage Efficiency


Dell EMC PowerMax Dell has recently acknowledged that the previous VMAX all-flash systems have achieved an average data reduction of 1.98:1 through the use of compression.1 The PowerMax uses a new and different compression algorithm, as well as an additional new post-ingestion data compression algorithm, so it is reasonable to expect that compression on the PowerMax should be improved over VMAX. Dell EMC, through its Future-Proof Loyalty Program, guarantees a 4:1 storage efficiency across its range of all-flash storage arrays, based on a combination of technologies that include compression, deduplication, thin provisioning, and snapshots. However, when considering data reduction alone, which includes compression and deduplication, Dell EMC is claiming an average 3:1 data reduction ratio for PowerMax. The PowerMax uses a 128KB comparison size for deduplication, which means that more granular data is unlikely to get deduplicated. With many applications writing at an 8KB block size, customers may not achieve significant data reduction from the PowerMax deduplication feature unless they are using very large block sizes (greater than 128KB). Additionally, Dell EMC notes that up to the most “active” 20% of data will not receive data reduction at all. Data reduction can be optionally enabled or disabled for each storage group.

Pure FlashArray//X

The Pure FlashArray//X is engineered to provide always-on data deduplication, compression, thin provisioning, encryption, snapshotting, ActiveCluster multi-site synchronous mirroring, Quality of Service (QoS) without any performance impact. The FlashArray//X delivers an average data reduction ratio (compression and deduplication only) of 5:1 data reduction. When combined with thin provisioning, the ratio increases to an over 10:1 average total efficiency rate. Pure Storage provides a simple, all-inclusive storage software licensing model.

Storage Management and Automation

Dell EMC PowerMax

Unisphere is the primary software set of tools for provisioning, managing, and monitoring PowerMax. It can be run on a dedicated server or run embedded directly in a container on the PowerMax OS hypervisor. In addition to extensive monitoring features, Unisphere provides a web-based interface to manage many system features, including managing user accounts, role-based access control, creating thin volumes, masking volumes, setting storage attributes, setting volume attributes, setting port flags, managing replication and backup operations, managing service levels, workload planning, performance troubleshooting, and much more. 5Users have a high degree of control over each of the features available on the PowerMax, and more often than not, users are required to provide their preferences for each of these features as part of the initial configuration and ongoing optimization of the system. To best understand the features and management of the PowerMax, Dell EMC has a 16-hour online learning training.

Pure FlashArray//X

Pure1® storage management software is a SaaS-based provisioning, management, and monitoring solution that integrates with Pure’s proactive support. It leverages machine-learning and predictive analytics to help advise customers on optimization and what-if situations, including capacity planning, and performance simulation. Pure1 capabilities are built on a global predictive intelligence engine called Pure1 Meta™ that leverages the accumulated data from the thousands of FlashArrays currently deployed. Pure1 Meta is the AI engine within Pure1 that provides the intelligence to manage, automate, and proactively support the FlashArray. Pure1 Meta collects more than a trillion telemetry data points of performance data per day. Part of the intelligence of Pure1 comes by way of its ability to recognize usage patterns. Pure1 identifies known patterns that may affect the optimal operations of FlashArrays, and notifies other FlashArrays with similar usage patterns of the concern. This way, customers are aware of potential impacts to their arrays and can proactively take preventive measures. Pure1 is browser-based and can be run without additional software to license, install, upgrade, manage, or change security policies. As new features become available, customers have immediate access to them. And with Pure1, organizations can manage, monitor, and analyze their storage from anywhere with any device, including mobile devices. Pure1 is consistent with Pure’s mantra of effortless operations, and results in a data platform that is selfmanaging and requires no tuning. In fact, for customers who prefer a command-line interface (CLI), there is usually no need to go beyond what is printed on the business card-sized Getting Started guide included with each FlashArray//X. The guide even includes all the necessary steps to establish a stretched multi-site replication cluster.

High Availability

Dell EMC PowerMax 

Although PowerMax is a new system, “PowerMAX is the next generation of the VMAX systems”, “It is the same scale-out architecture [as VMAX]”.3 The VMAX in turn evolved from Symmetrix DMX systems, which evolved from the original Symmetrix architecture. All of these previous generations of EMC systems established a proven degree of high-availability suitable for mission-critical application environments. Some of the world’s most demanding and critical application environments run in production on VMAX, and they almost always deploy multi-site replication with either SRDF or VPLEX to achieve the desired level of high-availability. Dell EMC claims that PowerMax arrays are architected for six-nines (99.9999%) availability. Either SRDF/Metro, or VPLEX/Metro are usually configured if the customer desires continuous operations in the event of an array failure.  Both SRDF/Metro licenses and VPLEX appliances and licenses, along with the associated professional services requires a significant upfront financial commitment from the customer, along with the associated software maintenance necessary to sustain the license. One of the promises of the Dell EMC Future-Proof Loyalty Program4 is “Never Worry Migrations”. PowerMax supports “Non-Disruptive Migration” (NDM) for migrations from VMAX to PowerMax. The NDM process flow is shown in Figure2 at right. In the same documentation, Dell EMC warns that “non-disruptive does not always equal no impact. NDM is designed to migrate a storage group (SG) at a time – not the entire system”.

Pure FlashArray//X 

Pure FlashArray//X and its previous generations have established a proven 99.9999% uptime suitable for missioncritical high-availability with a single system, without the requirement of remote replication to achieve that high degree of uptime. With Pure’s ActiveCluster, customers have the ability to extend that level of uptime even further. The Purity//FA operating environment that powers the FlashArray//X includes ActiveCluster at no additional cost. ActiveCluster provides the ability to link two different data center sites up to 150 miles apart in an active-active stretch cluster with transparent failover, zero recovery point objective (RPO) and zero recovery time objective (RTO). Pure’s ActiveCluster solution includes Pure1® Cloud Mediator, which is a software-based third entity that monitors the link between the two sites, and declares which site becomes the primary site, should the link fail. Pure1 Cloud Mediator runs in the Cloud, so no extra software and no extra hardware, and its associated maintenance, is needed. It can be used to provide rack-level active clustering inside a data center as well as linking separate data centers. A remote third data center can also be added for asynchronous replication, which is accessible and live for replication from both of the primary arrays in the ActiveCluster.


Dell EMC PowerMax

PowerMax engines or “Bricks” each have two redundant active-active controllers. The PowerMax 2000 supports up to two Bricks per system, and the PowerMax 8000 supports up to eight Bricks. Bricks own their own Disk Array Enclosures (DAEs) which house the NAND flash drives. The PowerMax 2000 can fit up to two Bricks in half a standard 19” rack, while the PowerMax 8000 can fit up to four Bricks in a single rack, and up to eight Bricks across two full racks taking up two full floor tiles. From whatever starting point a PowerMax system is initially configured, it can be grown to its maximum configuration by adding capacity and DAEs to existing Bricks, or in a “scale-out” fashion by keeping all of the existing Bricks and DAEs, and incrementally adding additional Bricks and DAEs, thus potentially increasing the footprint, as well as needing additional requirements for power and cooling. Historically, any upgrade between generations of VMAX systems required a full “forklift”-style upgrade and a complete data migration from the old system to the new system, as no two generations of engines/Bricks were able to coexist on 8the same system. Pure believes that the same situation has been carried forward to PowerMax, and we are unaware  of any claims by Dell to the contrary. Although the PowerMax 2000 is based upon the same architecture as the PowerMax 8000, it is not capable of delivering the same “scale-out” features as the PowerMax 8000, since it is limited to a system maximum of only two Bricks. In a single-brick configuration the PowerMax 2000 is simply a conventional dual-controller system, and in a two-Brick configuration it cannot “scale out” any further than its starting point, other than to possibly add capacity or cache.

Pure FlashArray//X 

Pure FlashArray//X uses a scale-up storage architecture which allows for a simpler and more flexible upgrade path. Pure’s Evergreen™ subscription to innovation makes it easy to scale from one model to the next (i.e. //M series → //X family) while keeping capacity in place, upgrading controllers non-disruptively, and even while in full-performance production mode. Due to Pure’s scale-up architecture, customers do not require forklift upgrades because Pure’s array controllers can be dynamically upgraded. Pure has a proven track record of not only enabling non-disruptive, data-in-place upgrades without performance impact or needing a data migration among systems of the same generation, but even between systems of different generations. For example, customers have upgraded from our FA-400 series, which leveraged two distinct controllers, to our current generation of integrated controller design – never having to move data, migrate data, or suffer any loss of accessibility or performance.

Leave a Reply

Your email address will not be published. Required fields are marked *

Latest Post

× How can I help you?