HPE Peer Copy: federating software-defined storage and all-flash arrays

HPE Peer Copy: federating software-defined storage and all-flash arrays

It’s not an either/or world. It’s about finding the right platform for each app, workload and service. Changing business and technology landscapes demand different IT requirements. This leads to different deployment models. Keeping the deployment models disparate is not an option.

HPE Peer Copy – an exciting new capability of our recently announced HPE RMC 4.0 software – enables you to centrally orchestrate a heterogeneous federated data center, bridging HPE StoreVirtual VSA software-defined storage with HPE 3PAR all-flash arrays and HPE StoreOnce data protection. 

Let’s take a closer look at the data mobility and data protection benefits and use cases delivered by bi-directional peer copy between 3PAR and StoreVirtual VSA.

Technology and business shifts demand different storage deployment models

A selection of storage deployment choice points are available to you as you manage the changing business landscape. To simplify things we’ve organised the world into two dimensions:

 

Performance versus cost

Do you need to deliver the best service-levels with flash optimized Tier 1 storage and the assured quality of service that comes from a shared storage system? The system defined storage view itself has shifted over the last several years with highly virtualized and scale-out systems displacing monolithic and dual-controller designs.

Or do you have significant amounts of business data and applications where you can lower costs by leveraging industry standard server hardware and the ability to co-locate applications, storage servers, and hypervisors on the same CPU and disks. Software-defined storage has taken off as processing power and server density has reached new levels.

Of course the answer is that you will probably need both, especially as you transform to a hybrid IT infrastructure.

Timing savings versus scalability

Deployment and consumption model discussions have turned to the use of Converged and Hyper-Converged systems as ‘easy-to-deploy-and-manage’ building blocks for IT. Use of these integrated systems is on the rise to accelerate time to value and free up data center resources.

However most workloads don’t scale with perfect ratios of compute and storage over time so there is still a need to grow your compute, networking, and storage resources flexibly to maintain efficiency.

Here again you will need both approaches as you transform to a hybrid IT infrastructure. The real key is to embrace these changing vectors and make decision points that don’t incrementally bring more complexity into what is already a very complex data center environment.

Deployment model coherence

Ideally you want to optimally utilize the resources and have the ability to seamlessly move workloads across these disparate storage tiers.

HPE Peer Copy – bi-directional replication between 3PAR service defined and StoreVirtual software defined storage

One step on the journey towards a heterogeneous federated data center - from software-defined to service defined all-flash to backup – is provided with the announcement of Recovery Manager Central 4.0.

 

The new HPE Peer copy feature within RMC 4.0 allows customers to bi-directionally replicate data volumes between StoreVirtual VSA and 3PAR arrays - eliminating the need to use host specific or hypervisor specific tools or expensive out of band network-based appliances.

HPE Peer Copy functionality is delivered at zero additional cost as RMC 4.0 ships as standard with every 3PAR 8000 and 20000 model as well as every StoreVirtual VSA.

How Peer Copy works

HPE Peer Copy – managed via the RMC 4.0 GUI - orchestrates and moves data from a source to a destination storage system. It automates the data copy process and makes the volume available for use after the copy without any need to convert the data.

 

When a data copy is triggered, RMC 4.0 creates a crash-consistent snapshot on the 3PAR or StoreVirtual VSA array. This is a simple process and decoupled from the actual movement of data, so the application only needs to be in backup mode for a brief instant. RMC then moves the snapshot data – from the source to the destination array. RMC has full access to the 3PAR and StoreVirtual VSA SnapDiff block-change detection technology. Therefore, it knows exactly what has changed since the last snapshot and only has to send those unique changed blocks at the array level from the source to the destination array. Multi-streaming enables even faster copy performance.

When the destination array receives the changed blocks from the source array, pointers are used to map previously written blocks along with the new ones so that a synthetic-full image copy of the data is always available for instant recovery whenever a restore is required.

During this process no host resources are used, virtually eliminating performance impact on the application.

Complementing snapshots and replication with backup

To deliver the comprehensive data protection that mission-critical applications require, you still need backups (in addition to snapshots and replication) as a second line of defence.

RMC 4.0 allows you to store self-contained, fully independent backup copies of your data on StoreOnce, protecting applications running on primary storage against storage platform outage as well as file loss or application corruption beyond the oldest snapshot. RMC 4.0 thus allows you to manage snapshots, replication and backup from a single GUI, enabling you to meet the many recovery SLAs that your business demands.

Check out my recent blog for an overview of why when it comes to best practice data protection, it should not be a question of ‘either-or’ but more a case of ‘where’ and ‘how’ to best use snapshots, replication and backup.

Use cases

Here are some potential use cases for the new RMC 4.0 bi-directional replication feature.

  • Edge to core protection: Helpful in ROBO scenarios, converged solutions such as a ProLiant server running StoreVirtual VSA running in the remote office can use a centralized 3PAR array in the main data center to house replicated volumes from remote sites. You can further reduce your risk exposure by using RMC 4.0 to backup these replicated volumes from 3PAR to StoreOnce. This reduces risk and benefits the customer with only needing a single storage and backup solution in the central office that can connect to software-defined storage solutions in the remote locations.
  • Dev/test to production data copy: The new replication capability can allow for more efficient data transfers in environments involving development and testing of applications and data prior to going into production. Customers can reduce application development costs by testing on software defined converged solutions in the dev, test lab and move to 3PAR in the production environment when ready.
  • Lower cost replication: The ability to use a software-defined storage solution such as StoreVirtual VSA can create a lower cost replication target option for 3PAR arrays within an existing server infrastructure, if purchasing a second 3PAR array does not fit within your budget. The secondary solution can still be located on-site or off as well as still providing business continuity and high availability of data.

Only HPE Peer Copy powered by RMC 4.0 can provide customers with a single management tool that federates storage providing integrated data protection and mobility for both software-defined and system-defined solutions.

Find Out More

HPE Peer Copy powered by RMC 4.0 - Infographic
HPE Peer Copy powered by RMC 4.0 – Chalk Talk Video
Replicating data from remote sites with HPE 3PAR and SDS – Solution Brief

Print
Simon Watkins

Simon Watkins  Simon Watkins

WW Product Marketing Manager, Backup, Recovery & Archive (BURA) HPE Storage

Other posts by Simon Watkins

Contact author

Name:
Email:
Subject:
Message:
x