QLogic 10000 Series Adapter Provides
Application Benets Through I/O Caching
QLogic FabricCache Caching Technology
Delivers Scalable Performance to
Enterprise Applications
The QLogic 10000 Series delivers scalable performance to
enterprise applications and reduces the performance gap
between these applications and their storage subsystems.
White Paper
KEY FINDINGS
The QLogic® FabricCache™ 10000 Series 8Gb Fibre Channel
Adapter enhances application performance by optimizing random
read-intensive workloads running in clustered, shared resource
environments. The QLogic 10000 Series technology integrates Fibre
Channel storage network connections, a ash memory interface, and
embedded processing. The technology delivers scalable performance
to enterprise applications and reduces the performance gap between
these applications and their storage subsystems.
QLogic’s caching technology in the 10000 Series Fibre Channel
Adapter delivers:
• Large application acceleration with modest cache congurations:
–In a single-server conguration, the adapter delivers 7.75× the
performance over having no cache
• Performance that scales directly with the server infrastructure:
–In a multiserver (four or more) conguration, the adapter
delivers up to 22× the performance over having no cache
• Remotely located cache performance difference of less than 10
percent compared to cache residing locally
EXECUTIVE SUMMARY
The growing disparity between the performance of CPU-based and
disk-based storage continues to grow. Applications that are Web 2.0,
mission-critical, I/O intensive, virtualized, and clustered continue to put an
additional burden on processors and slower storage, which lowers overall
application performance. These increased performance demands on the
enterprise data center require application performance improvements
using innovative solutions that can scale across the enterprise
infrastructure.
The QLogic 10000 Series Adapter technology delivers a unique solution
optimized to address the growing performance gap between what the
processor can compute and what the storage I/O subsystem can deliver.
With a simple adapter deployment model, the seamless approach
combines enterprise SAN connectivity with shared server-based I/O
caching. The QLogic 10000 Series Adapter provides dramatic and scalable
performance improvements to the widest range of enterprise applications.
This white paper provides an overview of the application performance
capability of the QLogic 10000 Series Adapter. A query workload from
the Transaction Processing Performance Council (TPC) Benchmark™
E (TPC-E) was used to generate database queries. TPC, an online
transaction processing (OLTP) workload, provides real-world, I/Ointensive transactions that simulate activities found in complex database
application environments. The QLogic 10000 Series meets the application
performance challenge and accelerates performance across the entire
enterprise.
SN0430935-00 Rev. D 04/14 1
QLogic 10000 Series Adapter Provides
Application Benets Through I/O Caching
White Paper
PROBLEM AND RESOLUTION
Increased server performance, higher virtual machine (VM) density,
advances in network bandwidth, and more demanding business
application workloads create a critical I/O performance imbalance
between servers, networks, and storage subsystems. Storage I/O is the
primary performance bottleneck for most virtualized and data-intensive
applications.
As application workloads and VM densities increase, so does the
pressure on these potential hotspots and the time required to access
critical application information. Slower storage response times result in
lower application performance, lost productivity, more frequent service
disruptions, reduced customer satisfaction, and, ultimately, a loss of
competitive advantage.
A new class of server-side storage acceleration is the latest innovation
in the market addressing this performance disparity. The idea is simple:
fast, reliable, solid-state Flash memory connected to the server brings
faster data access to the server’s CPU. Flash memory is highly available in
the market and promises to perform much faster than any rotational disk
under typical small, highly random, I/O enterprise workloads.
The QLogic 10000 Series Fibre Channel Adapter from QLogic shows
increased transaction throughput from 7 to 22times with OLTP database
workloads, while reducing application latency query (as measured
with TPC-E benchmarks). The 10000 Series Adapter is a PCIe-based
I/O device that provides integrated storage network (Fibre Channel)
connectivity, I/O caching technology, integrated ash memory, and the
embedded processing required to make management and caching tasks
transparent to the host server. The QLogic solution delivers the application
performance acceleration benets of server-based cache without the
limitations of solutions that require separate, server-based storage
management software and operating system (OS) lter drivers.
The key benets identied in this paper are the improvements by
the QLogic 10000 Series Adapter on the overall OLTP performance in
transactions per second (TPS). These improvements provide real-world
objective performance data to enterprise users. The primary objectives
are to optimize performance, maximize transactions, and increase the
value of the existing infrastructure, thereby reducing the costs of server,
management, and overall infrastructure.
• Moderate system and application execution time
• A balanced mixture of disk I/O and processor usage
SINGLE-SERVER QLOGIC 10000 SERIES ADAPTER DELIVERS
7.75 × PERFORMANCE
The single-server test setup shown in Figure 1 was congured with an
Intel® Xeon® Server connected to an HP® Enterprise Virtual Array (EVA)
storage array, through the QLogic 10000 Series Adapter and the QLogic
5800V/5802V Fibre Channel Switch. Benchmark Factory® for Databases
TPC-E runs were executed with cache levels varying in size from 0to
100percent of the actual SQL database size.
Figure 1. Single-Server Database Query Setup
Figure2 shows how performance improvements are directly related to
caching the application data closer to the server processor. The range
indicated by “1A” shows a signicant application performance available
with small congurations: 3×TPS at just 20percent cache size. The range
indicated by “1B” shows a large application acceleration available with
modest cache congurations: 7.75×TPS at 30percent cache size.
DATABASE WORKLOAD QUERIES WITH TPC-E
TPC-E, an OLTP workload, provides a mixture of read-only and updateintensive transactions that simulate the activities found in complex OLTP
application environments. TPC-E queries were used without publishing
any ofcial TPC-E benchmark results. The queries exercises a wide range
of system components associated with such environments, which are
characterized by these attributes:
• Simultaneous execution of multiple transaction types that span a
breadth of complexity
SN0430935-00 Rev. D 04/14 2
QLogic 10000 Series Adapter Provides
Application Benets Through I/O Caching
Figure 2: Single-Server Query TPS
MULTISERVER (4) QLOGIC 10000 SERIES ADAPTER DELIVERS
22 × PERFORMANCE
The four-server test setup shown in Figure3 was congured with four Intel
Xeon Servers connected to an HP EVA storage array through four QLogic
10000 Series Adapters and a QLogic 5800V/5802V Fibre Channel Switch.
TPC workload queries were executed in parallel with cache levels varying
in size from 0 to 100 percent of the SQL database size. The simultaneous
runs emulate a clustered application scenario, where numerous servers
are running I/O trafc to a single array.
White Paper
Figure 3. Multiple Server Database Query Setup
Figure 4 shows that in a four-server conguration, there is clearly a
very large aggregated TPS acceleration that is achieved with relatively
modest cache sizing. The range indicated by “2A” shows a signicant
application performance increase with small congurations: 3.5 × TPS at
just 20percent cache sized. The range indicated by “2B” shows that a
large application acceleration performance is available with modest cache
congurations and multiple servers: 22 × TPS at 30percent cache size.
SN0430935-00 Rev. D 04/14 3
QLogic 10000 Series Adapter Provides
Application Benets Through I/O Caching
White Paper
Figure 4: Four Server Database Query TPS
LOCAL CACHE VERSUS REMOTE CACHE EFFICIENCY OF QLOGIC 10000
SERIES ADAPTER REMOTE ACCESS
Figure5 identies the conguration testing the latency difference of local
versus remote cache (where a LUN cache is located in a 10000 Series
Adapter residing on another server). Test conclusions showed a nine
percent performance impact of slightly increased latency between local
versus remote database queries.
Figure 5: Local Versus Remote Test Conguration
As shown in Figure6, the performance benchmark measured 0percent
cache against 30percent local caching, and the test repeated with the
operation of caching at 30percent in the remote conguration. No
signicant difference was perceived between local cache access and
remote cache access.
Figure 6: Local Versus Remote Cache Efciency
TEST HARDWARE AND SOFTWARE
Table1 lists the hardware and Table2 lists the software used for database
workload queries.
Table 1. Hardware Used for Exchange Query
HardwareDescription
Dual 6-core Intel Xeon CPU
Host Server
Host Bus Adapter
Quantity of Host Bus Adapters
per Host
Fabric Switch
Storage ArrayHP EVA 6300
Drive Speed10K
Quantity of Drives on Array24
LUN Size100GB × 7
RAID TypeRAID 5
E5-2640 at 2.50GHz, 24logical
processors, 32GB RAM
QLogic FabricCache 10000 Series
8Gb Fibre Channel Adapter
1
QLogic 5800V/5802V 8Gbps Fibre
Channel Switch
SN0430935-00 Rev. D 04/14 4
QLogic 10000 Series Adapter Provides
Application Benets Through I/O Caching
White Paper
Table 2. Software Used for Exchange Query
SoftwareDescription
Operating System
I/O Load Generation Tool
Workload GeneratorTPC-E
Database TypeMicrosoft SQL Server
Database Size200GB
Database LUN Size1,000GB
Log LUN Size500GB
Microsoft® Windows Server®
2008 R2
Benchmark Factory for Databases
version 6.8
SUMMARY AND CONCLUSION
With the introduction of the FabricCache 10000 Series 8Gb Fibre Channel
Adapter, QLogic continues to be the industry leader in delivering highperformance I/O solutions to the data center and in solving enterprise
customer requirements.
Key performance ndings using the QLogic 10000 Series Adapter with
database applications:
• Performance gains are available to non-clustered applications—up to
7.75 × performance in a single-server conguration.
• Large performance gains are available for clustered applications—up to
22 × performance in a four-server conguration.
• The QLogic 10000 Series Adapter enables system performance to scale
with server infrastructure. Storage performance scales with server
compute and memory capacity.
• The QLogic 10000 Series Adapter provides efcient, remote access
support—less than 10 percent decrease between local congurations
and remote congurations.
Today’s virtualized and distributed workloads across the data center
demand faster I/O access. Existing solutions do not address the added
dimension of multiple servers acting together to run complex businesscritical applications, such as data warehousing, customer relationship
management (CRM), OLTP, enterprise resource planning (ERP), e-mail
servers, Web servers, reporting, analytics, nancial trading applications,
and cluster applications: Oracle® Real Application Clusters (RAC) and
IBM® DB2® pureScale®. Increased server performance, higher VM
density, advances in network bandwidth, and more demanding business
application workloads create a critical I/O performance imbalance between
servers, networks, and storage subsystems servicing these environments.
The QLogic 10000 Series 8Gb Fibre Channel Adapter provides best-inclass features and an unprecedented level of high performance, superior
scalability, and enhanced reliability. The unique caching solution of the
QLogic 10000 Series Adapter increases the server performance needed to
meet the escalating requirements of enterprise-class users. Performancehungry, response-time-sensitive applications demand increased
transactions and decreased response times. By locating application data
closer to the server CPU with the QLogic 10000 Series Adapter, the server
delivers lower latency for improved response times, as well as higher
throughput to feed those hungry servers.
The QLogic FabricCache 10000 Series Fibre Channel Adapter bridges
the transaction-based I/O performance gap and meets application
performance challenges by delivering scalable performance to enterprise
applications.
DISCLAIMER
Reasonable efforts have been made to ensure the validity and accuracy of these performance tests. QLogic Corporation is not liable for any error in this
published white paper or the results thereof. Variation in results may be a result of change in conguration or in the environment. QLogic specically
disclaims any warranty, expressed or implied, relating to the test results and their accuracy, analysis, completeness or quality.