Research Library

HPE Primera – Product Brief

Last updated June 18th, 2020 to update continuous async replication, and all NVMe SSDs supported. Concise and unbiased 5-page overview of the features and functionality of HPE Primera. Brief includes EvaluScale Reviewing Methodology. Download this free Product Brief now!


HPE Primera is a logical evolution of the HPE 3PAR StoreServ system with advances in the operating environment software, hardware architecture with multiple ASICs, and packaging. Data services offered with Primera are basically the same as with 3PAR. Notably, all advanced features are included in the base price with Primera. The file persona available with 3PAR is not offered with Primera.

HPE Primera is a scale-out block storage system that has three models. The architecture has up to four nodes with the internode communication handled by custom ASICs across a backplane. The larger models can scale from two to four nodes. The smallest model has two controllers, each with two processors (Intel Skylake) and four ASICs and 24 devices in the basic unit. The other models have two or four controllers and 48 devices. Systems scale larger with the addition of device enclosures. Model numbers prefixed with an “A” are all flash systems and with a “C” prefix can have HDDs installed.

Different from 3PAR is the Primera embedded operating environment which runs functions as independent “services” that can be updated and restarted without having to reboot the system. While not running Docker containers, the independent services are sometimes described as “containerized.” Each controller is independently running its own embedded environment with services.

Adaptive Flash Cache allows use of any installed SSDs as read cache extensions for DRAM.

HPE Adaptive Data Reduction includes deduplication of data as part of an overall data reduction collection along with zero detect, data compression, and data packing. Deduplication includes zero page elimination, thin provisioning along with snapshots and clones, and deduplication of data done inline. Deduplication is done on a 16KB granularity for selected volumes designated as TDVV (thin deduped virtual volumes). A Primera ASIC is used to perform the hash code of the 16KB block, a bit for bit comparison of data, and zero detect. A feature called Express Indexing does the initial table lookup of the hash code for detection of duplicate blocks.

Data compression uses functions of the Intel processor or an FPGA depending on the workload along with software to reduce data stored. Data Packing uses 16KB containers to store data independent of block boundaries for greater efficiency. With the Adaptive Data Reduction feature, HPE guarantees a 4:1 data reduction.

Both periodic and continuous asynchronous replication are supported in addition to synchronous remote replication.


The Primera system is an enterprise system that scales for use in both mid-range and high-end environments. HPE characterizes it as a Tier 0 system but it is a complete enterprise offering. There are many advanced features including the Adaptive Data Reduction and thin reclamation that maximize space utilization. Primera systems are targeted for block storage application requirements and storage consolidation needs.

  • Characteristics
    • Performance – HPE is very open about publishing performance data and has published vendor-tested numbers of 1.5M IOPs and 44 GB/s bandwidth. The SPC website should be consulted for the latest industry standard performance data.
    • Availability – Primera uses multiple controllers in a cluster arrangement and has the ability to failover in the case of a controller failure. Multiple redundancies of other components give the system high availability characteristics.
    • Replication for BC/DR – Remote copy for single or multi-volume copies (synchronous or asynchronous) and three-site DR and stretched cluster support is included as well as many-to-one data replication for consolidated data protection and archiving. Periodic and continuous asynchronous replication.
    • Deduplication, compression, and data packing to increase effective capacity.
    • Express Protect – direct backup to StoreOnce of snapshots
  • Applications
    • Primera is targeted at performance demanding applications that use block I/O in the enterprise space.
    • The features of snapshot and remote replication are required by enterprise applications and the Primera system with these advanced features included make it a good choice for those applications.
  • System environments
    • HPE Primera is a block storage system accessed using FC and iSCSI or NVMeoF in the future with standard drivers on host systems. There is a wide range of host operating systems supported.
  • Deployment and Administration
    • A new browser-based GUI on the Primera allows customers to install the system themselves in under 20 minutes according to HPE. Upgrades can be performed in five minutes without disruption.
    • To manage more than one system requires the separate management server – SSMC (StoreServ Management Console)
    • An internal analytics engine is used to monitor and adapt the Primera operation to optimize performance and availability.
  • InfoSight – SaaS analytics on storage system data
    • VM visibility
    • SRM type functions
    • Predictive failure and proactive service by HPE support

Evaluator Group EvaluScale™: HPE Primera – SAN Storage

Evaluator Group product review methodology “EvaluScale” assesses each product within a specific technology area.  The definitions of the criteria and explanations of how products are reviewed can be found in the  Evaluator Series Evaluation Guide. Download the product brief now to view the Evaluscale for this product.

Evaluator Group Opinion: Differentiating elements for HPE Primera

Primera is a logical evolution of 3PAR and HPE has not only advanced the product with introduction of technology and design updates, they have greatly improved the administration and added internal AI for optimization. The advances are needed to continue in a very competitive area. Not offering the file persona is also a very good decision since most customers in the high end of the enterprise market usually have a specific system for file and/or object storage.

Continuing with the advanced features is smart as well. These are expected capabilities, extensively used, and highly valuable. The maturity means there is low risk, and for those customers that have 3PAR, there will be no operational practice changes.
The scaling from two to four nodes will cover needs for enterprises as well. Using NVMe with flash and Optane devices yields tremendous performance improvements over earlier generations and the hardware intensive architecture with custom ASICs gives HPE a performance advantage.

Continuing advancement with InfoSight will increase the resiliency of the Primera systems and contribute to validating the 100% availability guarantee that will give a 20% of purchase price credit towards a future purchase or upgrade if not maintained. Management through the dashboards is made much simpler for a single system.

The Primera systems will eventually replace the 3PAR systems over time as a much improved enterprise storage system.


Register or login to download the free Product Brief with Evaluscale Product Review

Forgot your password? Reset it here.