Document # INDUSTRY2013002, July 2013 Page 2
Most medium and large sized IT organizations have deployed several generation of virtualized servers,
becoming more comfortable with the performance and reliability with each deployment. As IT organizations
started to increase VM density, they hit the limits of Hyper-V software and server memory, CPU, and I/O.
A new VM Engine is now available and this documents describes how it can help IT organizations maximize
use of their servers running Hyper-V in Windows Server 2012.
Best Practices in Hyper-V…………...……………..…………………………………………………...………..…3
Power Needed for More VMs………………………………………………..…………………………………….4
A New VM Engine…………………………………………………...……………………..……………..…………...5
A New VM Chassis ………………………………………..…………………………………………………………....6
Virtualized I/O………………………………....……………………………………….………………...……………..7
Racing Exhaust Systems: 16Gb FC, 10GbE & Converged Networking………..………….…..…..8
New Hyper-V Performance …………………………….………………………..………………………..……….9
Performance with 16Gb Fibre Channel………………..………………………...………….……………...10
Live VM Migration…..……………………………….…………………………………….…………..…………….11
Live Storage Migration………………….…………..…………….………………..…………….………………..12
Lowering the Cost of VM I/O………………………...………………………..…………………………………13
Low-Latency Connectivity……………………………………………………………………………….……...…14
Scalability with 16Gb Fibre Channel……………………………………………….………………………….15
Accelerating App Performance………………………………..……………………………………………...…16
The Bottom Line……….…………………………………………..……………………………………………….….17
Resources……….…………………………………………..……………………………………………….……………18
Harnessing the Power
In a survey conducted by IT Brand, IT professionals said the
average number of VMs per server would almost double in the
next 24 months.
VMs Per Server
Document # INDUSTRY2013002, July 2013 Page 3
Best Practices in Hyper-V
No hardware resource is more important to overall performance
than memory. Plan to ensure each VM has the memory it needs,
but without wasting memory in the process. Memory
Start with Planning
When planning a Hyper-V installation, it is important to take into account the new capabilities of Hyper-V in
Windows Server 2012. Windows Server 2012 has added significantly to the scalability of Hyper-V. For
datacenters virtualizing Tier-1 applications, the critical scalability enhancement is the ability to have up to
1TB of memory and 64 virtual CPU cores per VM. This will ensure almost all Tier-1 applications should
perform well in a Microsoft Hyper-V environment.
However, these new capabilities bring new complexities, and with them the need to plan new datacenter
architectures. This not only includes planning the deployment for today’s needs, but also thoroughly
investigating evolution strategies for applications before bolting down racks and filling them with servers.
Planning which applications are going to run on your virtualized servers is the first step in understanding
your needs.
From there, it is critical to define server integration points with existing resources (likely core switching
and storage resources), and how these will be affected by the evolution of existing resources.
After that, planning your approach to Live Migration and capacity growth over the lifetime of your new
infrastructure will help you scope internal I/O requirements appropriately.
Finally, determining whether to utilize converged networks or not, and what I/O performance you need, will
enable you to intelligently discuss your I/O and networking options with your SAN/LAN equipment providers.
These steps will help you ensure success when virtualizing your Tier-1 applications.
To fully optimize virtualized data centers, servers need maximum I/O capacity to support high input output
operation rates and high bandwidth applications. Increased bandwidth is also needed for server
virtualization, which aggregates I/O from multiple virtual machines (VMs) to the host’s data path. This next-
generation combination takes full advantage of new features that are described in detail in this planning
guide. Read on to discover how QLogic can increase your infrastructure ROI and overall competitiveness.
Document # INDUSTRY2013002, July 2013 Page 4
The Great VM Migration
With many server admins working on their 3rd and
4th generation of virtualized servers, the focus has
changed from interoperability and learning the
behavior of Hyper-V, to increasing VM density
(#VMs/Physical Host Server). With the availability of
servers based on Intel’s E5 processors (multi-core,
768GB of RAM, PCI Express Gen3) and the
combination of new features within Hyper-V, a new,
game-changing compute platform was introduced.
This new platform allows for new levels of VM
density and for the first time Tier-1 applications that
previously required dedicated server hardware can
now run on virtual servers, achieving improved
performance, scalability and efficiency.
While Hyper-V and E5-based servers are seeing
significant deployments in many enterprise
datacenters, the I/O and network infrastructure to
support these new technologies lags far behind. In a
survey conducted by IT Brand Pulse, IT professionals
said the average number of VMs per server would
almost double in the next 24 months. Approximately
25% of IT professionals surveyed also said what they
need most to increase the VM density is more I/O
bandwidth. The purpose of this industry brief is to
provide a planning guide to help enterprises deploy
Tier-1 applications with adequate bandwidth in a
dense Microsoft Hyper-V Server 2012 virtualization
environment.
Power Needed for more VMs
Approximately 25% of IT professionals surveyed said what they
need most to increase the density of VMs per server is more I/O
bandwidth.
VM Density
IT Brand Pulse
The average number of VMs per server in
my environment:
What I need most to increase the density of
VMs per physical servers is more:
IT Brand Pulse
Document # INDUSTRY2013002, July 2013 Page 5
A New VM Engine
2 more cores, 8MB more of cache, 6 more DIMMs of faster DDR3-
1600 memory increasing to 768GB, double the I/O bandwidth with
PCIe 3.0, and more Intel QuickPath links between processors. Xeon E5
Xeon E5 offers double the I/O bandwidth to 10GbE NICs, 10GbE/16Gb FC CNAs and 16Gb FC HBAs.
The Intel Xeon E5 Platform
The introduction of the Intel® Xeon® E5 family of processors responds to the call for more virtual server re-
sources with 2 more cores, 8MB more of cache, 6 more DIMMs of faster DDR3-1600 memory. Increasing
the total to 8 cores, 768GB of RAM, and doubling the I/O bandwidth with PCIe 3.0.
Intel’s new Xeon E5 promises a significant increase in server I/O by enabling full- bandwidth, four-port
10GbE server adapters as well as dual-port 16Gb FC server adapter support, addressing the VM density issue
with a substantial increase in I/O bandwidth that host servers require.
Intel Xeon E5 Platform
Document # INDUSTRY2013002, July 2013 Page 6
Windows Server 2012 Hyper-V
The newest release of Windows Server 2012 Hyper-V delivers a
high-performance VM chassis harnessing the new I/O
capabilities of 16Gb Fibre Channel (FC) Storage Networking and
10GbE Data Networking. Several new features of Windows Server 2012 Hyper-V are highlighted below:
vCPU—Virtual machines can now have up to 64 virtual CPUs (vCPUs) and 1TB of virtual RAM (vRAM) allowing Tier-
1 application to be virtualized and new levels of VM density to be reached.
Virtual Fibre Channel (Virtual FC)—Hyper-V now enables VM workloads to access FC SANs by provisioning virtual
FC ports with a standard Worldwide Name (WWN) within the guest OS.
Live Migration—Virtual FC also enables Live Migration of VMs across Hyper-V hosts while maintaining FC
connectivity. Two WWNs are configured and maintained for each virtual FC adapter.
Live Storage Migration—A VM’s Virtual Hard Disk (VHDX) storage can now be migrated without shutting down the
VM. The operation copies data from source storage device to a target via a FC or similar interconnect.
Multipath I/O (MPIO)— Hyper-V now extends MPIO capability to VMs ensuring fault-tolerant FC connectivity for
delivering High Availability and Resiliency to virtualized workloads
16Gb Fibre Channel—To help maximize efficiency of Live Migration and Storage Motion, Hyper-V includes support
for 16Gb Fibre Channel, the fastest storage interconnect available today.
10GbE and SR-IOV— Allows 10GbE NIC to appear as multiple virtual devices that can optimize I/O performance by
providing direct I/O for individual virtual machines.
From a storage planning perspective, when comparing Windows Server 2012 to previous version, there are
two specifications which stand-out: the amount of memory/VM (1TB), and the amount of active VMs per
machine (1,024).
Using today’s storage usage, a petabyte of storage could be needed to support 1024 VMs. While this
scenario is unlikely for at least a few years, running 100 VMs with 512GB of virtual memory each on a single
server (which would require 52TB of storage for the memory contents alone) is very foreseeable.
The ability to provide high-performance storage is critical for a high-density or Tier-1 virtualization strategy.
The new storage tools in Hyper-V that we will cover later in this paper (virtual Fibre Channel, offloaded data
transfer, and the new virtual hard disk format) can positively impact performance in these environments.
A New VM Chassis
In Hyper-V, Virtual Fibre Channel SAN connectivity on a
per VM basis.
Virtual Fibre Channel
Document # INDUSTRY2013002, July 2013 Page 7
Picking the Right I/O Pieces, and Making Them Work Together
Tier-1 applications are uniquely demanding in many dimensions. Their needs with respect to CPU power,
memory footprint, high availability/failover, resiliency and responsiveness to outside stimuli is typically
unmatched within the enterprise. Moreover, Tier-1 applications also tend to be tightly integrated with other
applications and resources within the enterprise. Because of this, virtualizing a Tier-1 application requires
rigorous planning of the I/O strategy. There are five steps to this:
Identify the I/O fabrics that the Tier-1 applications will use (it may very well be “all of them”).
Quantify the data flows for each fabric when the application was operating on a standalone system.
Estimate Live Migration I/O needs for failovers and evolution. Note that most Live Migration traffic will
be storage I/O; if the data stays within one external array during the Live Migration, Microsoft’s ODX
capability can significantly reduce the I/O traffic.
Determine your primary and secondary I/O paths for multi-pathing on all of your networks.
Determine QoS levels for the Tier-1 apps.
One simplifying option available is to utilize converged networking adapters that can function on both FC and
Ethernet/ FCoE networks. The QLogic QLE2672 is an example of such an adapter; it can be reconfigured in
the field to operate on 16Gb FC or 10Gb FCoE/ Ethernet networks.
Virtualized I/O
Business Critical Applications such as ERP, CRM, eCommerce
and Email need high-performance and high availability I/O
infrastructure to meet business SLAs. Tier-1 Apps
Networking Considerations When Virtualizing Tier-1 Applications
Document # INDUSTRY2013002, July 2013 Page 8
QLogic Server Adapters
Windows Server 2012 Hyper-V and Xeon E5 streamline the process of moving VMs and their associated
storage with Live Migration and Live Storage Migration and support low-latency networking traffic with SR-
IOV. Moving terabytes of data across the virtual machines and migration virtual servers requires low-latency,
high performance I/O adapters. QLogic offers a family of server adapters for 16Gb Fibre Channel , 10GbE or
converged network connectivity provide the bandwidth for increased Virtual Machine (VM) scalability and to
power Tier-1 application workloads.
Racing Exhaust Systems
The latest generation of CNAs from QLogic support Ethernet LANs, NAS,
iSCSI SANs and FCoE SANs, as well as native Fibre Channel SANs CNA
QLogic Server Adapters
2600 Series
3200 Series
8300 Series
Server Adapter
Description Fibre Channel HBA Intelligent Ethernet Adapter
Converged Network Adapter
Speed 16Gbps 10Gbps 10Gbps
Protocols FC TCP/IP LAN and NAS, iSCSI SAN
TCP/IP LAN and NAS, iSCSI SAN, FCoE SAN
Use in virtualized server
Highest performance Fibre Channel SAN connectivity for storage intensive applications
Consolidate multiple 1GbE server connections to LAN and NAS on one high-speed Ethernet wire
Consolidate server connections to LAN and SAN on one Ethernet wire
Document # INDUSTRY2013002, July 2013 Page 9
The size of a dynamically expanding VHD is as large as the data that is written to it. As more data is written to a dynamically
expanding VHD, the file increases to a maximum size. A differencing VHD is similar to a dynamically expanding VHD, but it contains
only the modified disk blocks of the associated parent VHD. Dynamically expanding VHDs are useful for testing environments
because there is less impact if you have to rebuild the VHD. For example, some of the tests performed for this report used multiple
dynamically expanding VHDs, each with a different Windows image. Fixed VHDs are recommended for production.
New Hyper-V Performance
VHDX A new Hyper-V virtual hard disk (VHD) format introduced with Windows
Server 2012 which increases storage capacity from 2TB to 64 TB
25% More Throughput with Windows Server 2012 Hyper-V
One of the most important features in Microsoft Windows 2012 Hyper-V for I/O performance is the VHDX
virtual hard disk format, which provides storage for the guest OS. Testing by Microsoft shows that VHDX
delivers nearly 25% better write throughput than VHD for both dynamically expanding and differencing disks.
VHDX Performance —1MB Sequential Writes
Document # INDUSTRY2013002, July 2013 Page 10
Testing with Iometer showed the performance of competitive products was identical at real world 4k, 8k and 16k block sizes.
However, the QLogic QLE2672 used up to 23% less CPU processing power to do the same work.
Performance with 16Gb FC
QLE2672 Testing by QLogic shows the QLE2672 16Gb Fibre Channel adapter
delivers high IOPS with Microsoft Windows Server 2012 with
significantly less CPU utilization than competitive products.
Real World Performance and CPU Efficiency
When used with a high-efficiency 16Gb Fibre Channel adapter, Hyper-V with VHDX can provide even larger
performance advantages. In testing performed by QLogic, the QLogic QLE2672 16Gb FC adapter delivered
the same IOPS performance using real-world 4KB & 8KB block sizes for dual-port adapters as the nearest
competitor—with 23% less CPU utilization. This frees CPU cycles for virtual machines and their workloads
which is critical for Tier-1 applications or in dense VM environments.
CPU % - Dual Port, 100% Reads IOPS - Dual Port, 100% Reads
IOPS Performance and CPU Utilization with 16Gb FC Adapters
Document # INDUSTRY2013002, July 2013 Page 11
ODX Data Copy Model Offloads Server
Microsoft’s Live Migration offers the ability to move a live
virtual machine from one physical server to another.
Included are some interesting storage-oriented capabilities
that provide added value if you use SANs (especially Fibre
Channel SANs). The first is the ODX Data Copy Model. On
compatible external storage arrays, ODX provides the ability
to move data between LUNs without involving the server.
For large data movements, this results in a huge
performance improvement.
Improved Live Migration with vFC
Hyper-V’s Virtual Fibre Channel (vFC) is a new capability
that augments Live Migration for those end users with Fibre
Channel SANs. By creating virtual FC HBAs natively within
Hyper-V, Microsoft simplifies migrations by moving the
adapter with the virtual machine. This eliminates the need
to re-configure network switches after a Live Migration.
Microsoft Hyper-V’s powerful migration
capabilities can provide even more utility if likely
failover and evolution paths in the private cloud
are planned into the framework. This is
especially true for migrations to resolve
hardware failures, which tend to be done under
considerable stress. Planning failover
migrations decreases the likelihood of negative
performance impacts that may ultimately have
to be undone later.
Live VM Migration
Windows Offloaded Data Transfer (ODX) in Windows Server 2012 directly
transfers data within or between compatible storage devices, bypassing
the host computer. ODX
This diagram shows a Live Migration utilizing vFC. The 2-port virtual HBA
ping-pongs from the first port to the second port, avoiding traffic disruption
during the live migration.
Traditional Data Copy Model ODX Data Copy Model
1. A user copies or moves a file or this occurs as part of a virtual machine migration.
2. Windows Server 2012 translates this transfer request
into an ODX and creates a token representing the data.
3. The token is copied between the source server and des-tination server.
4. The token is delivered to the storage array.
5. The storage array internally performs the copy or move
and provides status information to the user.
Document # INDUSTRY2013002, July 2013 Page 12
16Gb Fibre Channel Helps Close the Storage Migration Window
A process known as Live Storage Migration allows for a non-disruptive migration of a running VM disk files
between to different physical storage devices. This process allows for the virtual machine to remain running
with no need to take its workload offline to move the VM’s files to a different physical storage device.
Additional use cases for Live Storage Migration include migration of data to new storage arrays or larger
capacity, better performing LUNs. NPIV zoning and LUN masking must be properly configured to ensure the
VM and host server continue to have access to the storage after the migration is completed. Live Storage
Migration across a 16Gb Fibre Channel link can finish in half the time it takes a 8Gb Fibre Channel link. All
paths related to Live Storage Migration should be supported by high performance networks in order to
reduce the time it takes to evacuate storage safely to a new destination and resume normal operations.
Additionally, 10GbE links can replace 1GbE links to ensure proper bandwidth exists for Live Storage Migration
in Ethernet environments.
Live Storage Migration
A single port QLE2670 16Gb Fibre Channel HBA doubles throughput for a live storage migration.
Storage Live Migration at 16Gbps
This bandwidth intensive operation now enables virtual
machines and associated VHDX files to be migrated between
clusters that do not have a common set of storage.
Live Storage Migration
QLE2670 16Gb Fibre
Channel HBAs
QLE3240 10Gb
Intelligent Ethernet
Adapters
Document # INDUSTRY2013002, July 2013 Page 13
8300 Series CNAs Enable Convergence at the VM edge
QLogic Converged Network Adapter solutions leverage core technologies and expertise including the most established and proven driver stack in the industry. These adapters are designed for next-generation, virtualized, and unified data centers with powerful multiprocessor, multicore servers. Optimized to handle large numbers of virtual machines and support for VM aware network services with support for concurrent NIC, FCoE, and iSCSI traffic.
One 8300 series CNA can be configured for connectivity to an Ethernet network and to deliver storage networking via Fibre Channel over Ethernet simultaneously. Powerful iSCSI and FCoE hardware offloads improve system performance and advanced virtualization technologies are supported through secure SR-IOV or switch and OS agnostic NIC Partitioning (NPAR). Combine with QLogic’s Quality of Service (QoS) capability for consistent and guaranteed, application aware performance in dense VM environments.
Lowering the Cost of VM I/O
The Fibre Channel over Ethernet (FCoE) protocol allows Fibre Channel
traffic to run over Data Center Ethernet (DCE) for LAN and SAN
convergence on one wire. FCoE
For organizations maintaining a parallel LAN and SAN architecture all the way to the server adapter, QLogic offers the
QLE8300 Series of adapters supporting 10GbE LAN, NAS and iSCSI SAN traffic, as well as Fibre Channel traffic.
Network Convergence at the VM Server
Adapter & Fabric Convergence Adapter Convergence, Separate Fabrics
Both ports used for
LAN, NAS and SAN
traffic over Ethernet
Ethernet ToR Switch Ethernet ToR Switch FCoE ToR Switch
1 port used as FCoE CNA
1 port used as Ethernet NIC
Document # INDUSTRY2013002, July 2013 Page 14
8300 Series CNAs Offload the VM Kernel from Switching Virtual NICs
Single Root I/O Virtualization is a standard that allows one PCI Express (PCIe) adapter to be presented as
multiple separate logical devices to virtual machines for partitioning adapter bandwidth. The hypervisor
manages the Physical Function (PF) while the Virtual Functions (VFs) are exposed to the virtual machines. In
the hypervisor, SR-IOV capable network devices offer the benefits of direct I/O, which includes reduced
latency and reduced host CPU utilization. With SR-IOV, pass through functionality can be provided from a
single adapter to multiple virtual machines through Virtual Functions. To deploy SR-IOV today, an
organization needs to ensure a minimum level of infrastructure (server hardware and OS) support for SR-IOV.
In contrast, QLogic NPAR technology can similarly be used today without the minimum levels of
dependencies of SR-IOV.
Low-Latency Connectivity
Latency is the time between the start and completion of one action
measured in microseconds (µs) . Latency
With SR-IOV enabled on a 10GbE NIC, pass through functionality can be provided from a single adapter to
multiple virtual machines through Virtual Functions (VFs).
Implementing Pass-Through Functions with SR-IOV
8300 Series CNAs
Document # INDUSTRY2013002, July 2013 Page 15
In Transaction Intensive and Bandwidth Intensive Environments
For virtualized environments, the most critical measure of performance is the ability to scale as the number
of VMs and application workloads increase. In testing conducted by QLogic, the QLogic QLE2672 delivered
three times the transactions and double the bandwidth of 8Gb Fibre Channel Adapters. The QLE2672 also
demonstrated a 50% advantage over competitive products for read-only performance and 25% better mixed-
read-write performance. The superior performance of QLogic 16Gb Fibre Channel Adapters translates to
support for higher VM density and support for more demanding Tier-1 applications.
QLogic achieves superior performance by leveraging the advanced 16Gb Fibre Channel and PCIe® Gen3
specifications—while maintaining backwards compatibility with existing Fibre Channel networks. The unique
port-isolation architecture of the QLogic FlexSuite adapters ensures data integrity, security and deterministic
scalable performance to drive storage traffic at line rate across all ports. QoS enables IT teams to control and
prioritize traffic. And paired with adapter partitioning technology the QLE2672 can deliver capacity on
demand and multitenant feature requirements of highly virtualized environments.
More Virtual CPUs for scaling Tier-1 Apps
If you’re concerned about hosting tier-1 apps on VMs,
the argument about virtualizing tier-1 apps is over.
Even flagship enterprise applications such as
Microsoft SQL Server 2012 and Exchange 2010 have
adopted server virtualization as a best practice. In
fact the CPU, memory, storage and networking
requirements are well documented by Microsoft.
In the example on the right, a mission-critical OLTP
Workload running on a single SQL Server 2012 VM
demonstrates linearly increasing transactional
performance and reduced transaction response times
as the number of virtual CPUs assigned to the
workload are increased to the maximum of 64 now
supported on Hyper-V.
Scalability with 16Gb FC
Throughput Bandwidth refers to the maximum potential volume. Throughput
is the actual volume. Both are measured as the amount of data
transferred in a given time or megabytes per second (MBps).
The number of transactions processed per second and the
average response time were monitored as virtual CPUs were
increased from 4-64. The OLTP workload and concurrent user
counts remained constant.
Hyper-V Virtual CPU Scalability With OLTP Workloads
(Source: Microsoft )
Document # INDUSTRY2013002, July 2013 Page 16
10000 Series FabricCache Adapters Cache Hot VM Data
The 10000 Series is the industry's first caching SAN adapter. This new class of server-
based PCIe SSD/Fibre Channel HBAs uses the Fibre Channel network to cache and share
SAN metadata. Adding large caches to servers places the cache closest to the application
and in a position where it is insensitive to congestion.
An advantage to this approach is that PCIe flash based caching can be shared and
replicated in different servers for high-availability and for cache coherency across
migrating servers in a virtual machine cluster. With the FabricCache architecture, the new generation of PCIe
SSDs provide redundancy and fail-over for a new level of enterprise-class availability.
Accelerating App Performance
A QLogic architecture for sharing and replicating cache on a PCIe
SSD adapter in a SAN. FabricCache
The lightning fast SSD SLC flash from the 10000 Series FabricCache adapters is used to cache hot data stored on a FC SAN array. For
high availability, the cache LUNs from a FabricCache adapter in one server can fail-over to a FabricCache adapter in another server
and can also be used for cache coherency across migrating servers in a virtual machine cluster.
Shared PCIe SSD
Document # INDUSTRY2013002, July 2013 Page 17
The Bottom Line
The improvement factor for Memory per VM for Windows Server 2012
Hyper-V — addressing the biggest issue in scaling VMs. 16X
More VMs with Hyper-V, Xeon E5 and QLogic Server Adapters
Fabric-based networks are a fundamental requirement in supporting highly virtualized data centers. Fibre
Channel SANs are the nucleus of the next-generation Windows Server 2012 data center. If your goal is to
increase VM density, Windows Server 2012 Hyper-V, combined with the latest generation of servers based
on Intel Xeon E5 processors, and QLogic server adapters allow you to more than double the number of VMs
per server while enjoying the same level of performance. Virtualization features like Microsoft vFC and Fibre
Channel QoS from QLogic combine to deliver reliability, performance, and the flexibility necessary to manage
the complexity and risks associated with virtualization projects. Choosing to virtualize tier-1 data center
applications or increase virtualization densities with QLogic and Hyper-V will enable your businesses to
leverage the built-in architecture of both products to increase availability, improve agility, and overcome
scalability and performance concerns.
Hyper-V delivers improvements on all key virtualization metrics—making I/O performance critical.
Windows Server 2008 R2 Hyper-V 2012 Hyper-V Factor
Host
HW Logical Processors 64 LPs 320 LPs 5x
Physical Memory 1 TB 4 TB 4x
Virtual CPUs per Host 512 2048 4x
VM
Virtual CPUs per VM 4 64 16x
Memory per VM 64GB 1TB 16x
Active VMs per Host 384 1024 2.7x
Guest NUMA No Yes -
Cluster Max Nodes 16 64 4x
Max VMs 1,000 8,000 8x
Related Links
What’s New in Hyper-V—Platform
What’s New in Hyper-V—Networking
What’s New in Hyper-V— Virtual Fibre Channel Storage
What’s new in Hyper-V—Storage Migration
QLogic Fibre Channel Adapters
QLogic Converged Network Adapters
Acceleration for Microsoft SQL Servers
About the Authors
Rahul Shah, Director, IT Brand Pulse Labs
Rahul Shah has over 20 years of experience in senior engineering and product management
positions with semiconductor, storage networking and IP networking manufacturers including
QLogic and Lantronics. At IT Brand Pulse, Rahul is responsible for managing the delivery of
technical services ranging from hands-on testing to product launch collateral. You can contact
Rahul at [email protected].
Tim Lustig, Director of Corporate Marketing, QLogic Corporation
With over 18 years of experience in the storage networking industry, Lustig has authored
numerous papers and articles on all aspects of IT storage, and has been a featured speaker at
many industry conferences on a global basis. As the Director of Corporate Marketing at
QLogic, Lustig is responsible for corporate communications, , third party testing/validation,
outbound marketing activities and strategic product marketing directives of QLogic. His
responsibilities include customer research, evaluation of market conditions, press and Media
relations, social media and technical writing.
Resources