NetApp and Broadcom recently published a new NetApp® Verified Architecture: a best practice reference architecture that illustrates an optimally configured VMware installation on NetApp ONTAP® 9.6 and Broadcom technology (Emulex 32GB HBAs and Brocade 32GB FC switches). If you manage, administer, or purchase VMware-virtualized configurations or use VMware with SAN or NAS storage, read NVA-1136-Design. You’ll learn about the benefits of an ONTAP and Broadcom SAN that provides storage for your VMware environments, and you’ll see how you can modernize them to:

  • Optimize performance, availability, and flexibility
  • Protect and guarantee workloads in shared environments
  • Use NetApp encryption offerings to secure and protect your data with no impact on performance

This NetApp Verified Architecture document, titled “NetApp and Broadcom Modern SAN Cloud-Connected Flash Solution,” explains the advantages of using a NetApp and Broadcom SAN to provide VMware storage for datastores and raw device mappings (RDMs). The document also describes tests that you can easily reproduce. By reading about the test data, configurations, and other details, you can re-create the test environment and obtain the excellent performance results that we observed.

 

The document also illustrates how to gain other benefits from the solution. For example, it describes:

  • How to setup data protection and backup solutions. You can apply hyperscalers or an on-premises hybrid cloud approach by using the NetApp StorageGRID® (cloud-connected or local) object storage appliance.
  • NetApp hypervisor plug-ins like NetApp Virtual Storage Console (VSC), a VMware plug-in that enables storage provisioning directly from the vSphere management console.
  • NetApp SnapCenter® management and its ability to manage data protection and disaster recovery relationships for all of your VMware environments.
  • Future proofing provided by NetApp’s industry-leading, innovative NVMe offerings and its NVMe over Fibre Channel (NVMe/FC) support. By migrating from SCSI-based protocols like FC Protocol and iSCSI to NVMe/FC, you can increase performance and reduce latency, which can help extend the service life of your storage arrays.

The real beauty of this solution is that, by using this reference architecture, you can achieve the value proposition and performance gains of ONTAP today. The document lays out configuration details and best practices you can easily use to build your own environment and achieve the same best-in-class performance, availability, scalability, and storage efficiency. The architecture can optimally host any applications or environments you deploy. And because NetApp engineering is constantly testing and analyzing the performance, reliability, and scalability characteristics of ONTAP components, you’ll see performance increases and new capabilities with each ONTAP upgrade, at no additional cost. For example, upgrades from ONTAP 9.2 to 9.3 or later typically saw an iSCSI performance increase of up to 40% because engineering rewrote the ONTAP iSCSI target. A stack rewrite between ONTAP 8.3 and 9 created similar FC Protocol performance gains. In addition to improving performance, NetApp engineering is almost fanatical about hardening ONTAP and cluster resiliency and making fault recovery graceful and nondisruptive.

 

Although ONTAP supports NVMe/FC today, not all operating systems have updated their initiator stack to add NVMe support yet. But we expect major OS vendors to announce NVMe support in their next operating system releases. For instance, it’s likely that VMware will announce NVMe support in ESX.next at VMworld this year.

 

Once VMware adds NVMe support, customers that have adopted the configuration (or similar configurations) would, with an ESXi upgrade, be able to improve performance and reduce latency substantially by migrating existing FC Protocol workloads to NVMe. By migrating to NVMe, you’re likely to see these benefits:

  • Performance can improve significantly. In NetApp testing, we’ve often seen IOPS increase 5060% when we moved from FC Protocol to NVMe. The increase is due to the massive protocol efficiencies designed into NVMe. These efficiencies include removal of software locks, reduction in context switching, and major parallelization improvements enabled by the huge number of queues and queue depths supported in NVMe. For more information about NVMe and its architectural improvements, review TR-4684.
  • NetApp and other industry tests show that NVMe reduces host processor utilization. Reduced CPU utilization can mean a reduction in power, cooling requirements, data center tiles, number of ESXi servers necessary to support specific workloads, administrative effort, and licensing.
  • Hardware lifespans might be extended. By migrating to the much more efficient NVMe/FC protocol, existing hardware can achieve 50% more IOPS. You might be able to squeeze more life out of existing infrastructure investments before performance pressures require upgrades in the environment.

For more information, check out these reports:

Contact your NetApp account team to get an ONTAP briefing that covers ONTAP SAN, running virtualized infrastructures on ONTAP, ONTAP NVMe, and other areas where NetApp can provide highly efficient, scalable solutions. These solutions can grow with you and can encompass on-premises environments, hybrid clouds, and cloudmobile solutions, all with industry-leading data protection and storage efficiency.

Michael Peppers

Michael Peppers is a technical marketing engineer (TME) with a focus on SAN, FlexArray/V-Series, QoS, FLI, and NVMe in ONTAP. His TME mission is to broaden and deepen the SAN, FlexArray & related knowledge of customers, partners, and coworkers, and increase the scope of NetApp's SAN & FlexArray solutions. Prior to becoming a TME Mike tested ONTAP SAN Interoperability in engineering quality assurance, before that, was an Escalations Engineer in the NetApp Technical Support Center. Before joining NetApp in 2005, Mike was a Networking and Systems administrator for several of US-based corporations.