SlideShare a Scribd company logo
RDP 7.0 + IPQ TEST REPORTPrepared by IPeak Networks Incorporated22 October 2010<br />Executive Summary<br />Test Data Highlights – IPQ Improvements<br />Interpretation<br />WAN Scenarios<br />Satellite Scenarios<br />Retransmission Analyses<br />Conclusions<br />Full Test Results<br />IPQ is a very effective RDP optimizer for lossy networks<br />When measured in terms of bandwidth utilization IPQ improved RDP 7.0 performance in all tests where packet loss is greater than zero.<br />High bandwidth applications benefit greatly with up to 235% improvement in data transfer speeds to optimize application performance under lossy network conditions and provide corresponding improvements in the user experience, i.e., far fewer delays and a more fluid and productive experience.<br />Packet Loss and High Bandwidth Applications (e.g., PPT, Flash) <br />Loss and latency cause TCP “clamping” and reduced bandwidth. <br />Reduced bandwidth is experienced as reduced application performance.<br />IPQ- protects against loss<br />- prevents TCP clamping<br />- increases bandwidth, and<br />- restores application performance <br />Low bandwidth applications also benefit with the near complete elimination of TCP retransmissions and a significant reduction in the user’s experience of lag.<br />Packet Loss and Low Bandwidth Applications (e.g., Word) <br />Loss and latency triggers TCP retransmissions<br />TCP retransmissions are experienced as delayed application response or “lag”<br />IPQ- protects against loss<br />- reduces retransmissions, and<br />- reduces lag<br />IPQ IMPROVEMENTS - All Test Scenarios<br />WAN SCENARIOS<br />WAN and PowerPoint - Significant Reduction in Test Time, Increase in Average Speed<br />In these tests, packet loss caused significant delays, uneven application performance, and an unsatisfactory user experience.  With IPQ protection against that packet loss, the delays were all but eliminated, the application performance was smooth and steady, and the user experience was comparable to the experience over the LAN.<br />Test Times Reduced<br />Over the WAN with 3% loss, this test required an additional 00:08:30 to complete when compared to the Calibration baseline test time of 00:41:01 [See Full Test Results].  With IPQ protection against that packet loss, 00:03:39 of that time was recovered.<br />The IPQ Improvement when the WAN suffers 5% packet loss is even more dramatic.  At 5% packet loss, this test required an additional 00:27:36 to complete when compared to the baseline.  With IPQ protection against packet loss, 00:21: 24 of that additional time was recovered.<br />With zero packet loss over the WAN, this test ran in 00:44:44.  With 5% packet loss over the WAN, this test ran in 01:08:37.  With IPQ protection against packet loss, this test ran in 00:47:13.  With IPQ, the impact of packet loss on the test completion time was practically eliminated.<br />Speed (bps) Increased<br />IPQ improved average data transfer speeds in these tests.  With IPQ Protection against 3% packet loss, average speeds were improved by more than 22%.  At 5% packet loss, IPQ protection enabled an average data transfer speed improvement of more than 67%.<br />WAN SCENARIOS (con’t)<br />WAN and Word – Notable Reduction in Lag<br />Test Times, Speeds Unchanged<br />The measurements of test times, throughput, and average data transfer speeds in this test scenario confirmed that low bandwidth applications such as Word are not impacted by the bandwidth constraints induced by the interaction of packet loss and TCP.  Test results with/without packet loss and with/without IPQ were not materially different.<br />Retransmissions Reduced, User Experience Enhanced<br />However, packet loss did cause delays in the application response times resulting in a notable deterioration in the user experience.  With IPQ protection, the number of TCP retransmissions required to complete deliver user input to the application and return application response to the user was greatly reduced.  I.e., the “lag” experienced by the user was significantly reduced.  See the separate Retransmission Analyses for more. <br />WAN SCENARIOS (con’t)<br />WAN and Flash – Playout Times Corrected, Data Transfer Speeds Improved<br />Like PowerPoint, Flash is a high bandwidth application and as with the WAN PPTx tests packet loss caused significant delays, uneven and even un-useable application performance, and an unacceptable user experience in this test scenario.  With IPQ protection against the packet loss, delays were greatly reduced, the application performance was smooth and steady, and the user experience was comparable to the experience over the LAN i.e., Flash was useable.<br />Test Times Reduced<br />Over the WAN with 5% loss, this test ran in 00:14:34, almost 6X the actual 00:02:30 runtime on the LAN [See Full Test Results].  With IPQ protection against the packet loss, this test time was reduced by more than half.<br />Speed (bps) Increased <br />IPQ improved average data transfer speeds in these tests.  With IPQ Protection against 3% packet loss, average speeds were improved by more than 10%.  At 5% packet loss, IPQ protection enabled an average data transfer speed improvement of more than 61%.<br />SATELLITE CONNECTION SCENARIOS<br />Satellite Connection and PowerPoint – Extreme Reduction in Test Time, Extreme Increase in Average Speed<br />As with the WAN PPTx tests, packet loss caused significant delays, uneven application performance, and an unsatisfactory user experience in the Satellite PPTx tests.  With IPQ protection against that packet loss, the delays were reduced dramatically, the application performance was far more fluid steady, and the user experience was much closer to the experience over the LAN.<br />Test Times Reduced<br />Over a Satellite Connection with 3% loss, this test required an additional 01:07:36 to complete when compared to the Calibration baseline test time of 00:41:01 [See Full Test Results].  With IPQ protection against that packet loss, 00:52:46 of that time was recovered.<br />The IPQ Improvement when the satellite Connection suffers 5% packet loss is even more dramatic.  At 5% packet loss, this test required an additional 01:25:03 to complete when compared to the baseline.  With IPQ protection against packet loss, 01:08:18 of that additional time was recovered.<br />With zero packet loss over the Satellite Connection, this test ran in 00:55:27.  With 5% packet loss over the Satellite Connection, this test ran in 02:06:04.  With IPQ protection against packet loss, this test ran in 01:08:18.  With IPQ, the impact of packet loss on the test completion time was dramatically reduced.<br />Speed (bps) Increased<br />IPQ improved average data transfer speeds in these tests by as much as 2.3X.  With IPQ Protection against 3% packet loss, average speeds were improved by more than 97%.  At 5% packet loss, IPQ protection enabled an average data transfer speed improvement of more than 234%.<br />SATELLITE CONNECTION SCENARIOS (con’t)<br />Satellite Connection and Word – Notable Reduction in Lag<br />Test Times, Speeds Unchanged<br />The measurements of test times, throughput, and average data transfer speeds in this scenario confirm that low bandwidth applications such as Word are not impacted by the bandwidth constraints induced by the interaction of packet loss and TCP.  Test results with/without packet loss and with/without IPQ were not materially different.<br />Retransmissions Reduced, User Experience Enhanced<br />However, packet loss did cause delays in the application response times resulting in a notable deterioration in the user experience.  With IPQ protection, the number of TCP retransmissions required to deliver user input to the application and return application response to the user was greatly reduced.  I.e., the “lag” experienced by the user was significantly reduced.  See the separate Retransmission Analyses for more.<br />SATELLITE CONNECTION SCENARIOS (con’t)<br />Satellite Connection and Flash – Playout Times Improved, Data Transfer Speeds Doubled<br />Like PowerPoint, Flash is a high bandwidth application and as with the Satellite Connection PPTx tests packet loss caused frequent, prolonged delays, and a largely un-useable user experience in this test scenario.  With IPQ protection against that packet loss, the delays were greatly reduced, and the application performance was improved as was the user experience.<br />Test Times Reduced<br />Over the Satellite Connection with 5% loss, the Flash test completed in 00:37:26, many times the actual Flash running time of 00:02:30 on the LAN [See Full Test Results].  With IPQ protection against packet loss, the test time was reduced by a full 00:22:16 bringing Flash over a Satellite Connection into the realm of possibility.<br />Speed (bps) Increased<br />IPQ improved average data transfer speeds in these tests.  With IPQ Protection against 3% packet loss, average speeds were improved by more than 121%.  At 5% packet loss, IPQ protection enabled an average data transfer speed improvement of more than 118%.<br />RETRANSMISSION ANALYSES<br />WAN and Word – Dramatic Reductions in Retransmissions and Lost Keystrokes<br />The analysis of the retransmissions with and without IPQ was prompted by the observation of high levels of lag or delays in the application response to user input, i.e. to the typing and scrolling activities.  Keyboard input that fails to reach the virtual machine and application due to packet loss requires retransmission.  The application response is delayed until the retransmission is made successfully and the keystroke is communicated to the application.  Similarly, the application response to the keystroke is also subject to potential loss, also requiring retransmission from application to user.  Whether lost enroute to or from the application or the user – or both – the loss and retransmission manifest as lag.<br />TCP retransmissions are at once symptomatic of packet loss and the programmatic response to packet loss.  IPeak Networks undertook this additional retransmission analysis to provide a set of underlying measurements that may be correlated to the user experience of lag.<br />Retransmissions and Lost Keystrokes Reduced<br />At 3% packet loss, while typing and scrolling over the WAN, retransmissions accounted for more than 11.47% of the data transferred.  In effect, approximately one of every nine keystrokes was lost due to packet loss and required retransmission.  With IPQ protection against the packet loss, the retransmissions accounted for 0.57% of the data transferred, a net reduction in retransmissions equal to more 95%.  With IPQ, the number of keystrokes lost due to packet loss was reduced to one in 200.<br />When typing and scrolling at 5% packet loss, retransmissions accounted for 14.89% of the data transferred.  IPQ protection reduced the retransmissions in this test by almost 90% and the lost keystrokes were reduced from about one in six to fewer than one in sixty.<br />RETRANSMISSION ANALYSES (con’t)<br />Satellite Connection and Word – Dramatic Reductions in Retransmissions and Lost Keystrokes<br />The analysis of the retransmissions with and without IPQ was prompted by the observation of high levels of lag or delays in the application response to user input, i.e. to the typing and scrolling activities.  Keyboard input that fails to reach the virtual machine and application due to packet loss requires retransmission.  The application response is delayed until the retransmission is made successfully and the keystroke is communicated to the application.  Similarly, the application response to the keystroke is also subject to potential loss, also requiring retransmission from application to user.  Whether lost enroute to or from the application or the user – or both – the loss and retransmission manifest as lag.<br />TCP retransmissions are at once symptomatic of packet loss and the programmatic response to packet loss.  IPeak Networks undertook this additional retransmission analysis to provide a set of underlying measurements that may be correlated to the user experience of lag.<br />Retransmissions and Lost Keystrokes Reduced<br />Retransmissions represented 11.49% of the data transferred when typing and scrolling over a Satellite Connection with 3% packet loss.  One in nine keystrokes was lost.  With IPQ protection against packet loss, and retransmissions were reduced by almost 90 percent and lost keystrokes were reduced by from one in nine to one in 62.  At 5% packet loss, retransmissions were reduced by more than 80% and lost keystrokes were reduced from approximately one in 6 to fewer than one in forty.<br />CONCLUSIONS<br />The results of these tests confirm that packet loss and latency, both individually and in combination, have a seriously damaging effect on the performance of both high bandwidth and low bandwidth applications running in virtual desktops. It follows that it is desirable to make efforts to reduce latency and reduce packet loss in order to preserve applications performance and support a satisfactory user experience of the virtual desktop.<br />There is little that can be done to reduce network latency other than locating the data centers and remote desktop users in close physical proximity to one another, a strategy that is antithetical to the vision of a ubiquitous and even global deployment of virtualized computing.  It is therefore reasonable to focus on the performance gains and user experience benefits that may be realized through the use technologies that reduce packet loss.<br />The results of these tests further confirm that IPQ is an effective solution for the problem of network packet loss.  When measured in terms of reduced test time-to-completion, increased thoughput, increased average data transfer speeds, and reduced TCP retransmissions, IPQ proved to be very effective in optimizing RDP 7.0 performance in all tests where packet loss is greater than zero.<br />High bandwidth applications benefited greatly with up to 235% improvement in data transfer speeds helping to optimize application performance under lossy network conditions and provide corresponding improvements in the user experience, i.e., far fewer delays and a more fluid and productive experience.<br />Low bandwidth applications also benefit with the near elimination of TCP retransmissions, a vast reduction in the number of keystrokes lost as a function of packet loss, and a corresponding reduction in the user’s experience of lag.<br />-8204201109980<br />
IPQ and RDP Test Report
IPQ and RDP Test Report
IPQ and RDP Test Report
IPQ and RDP Test Report
IPQ and RDP Test Report
IPQ and RDP Test Report
IPQ and RDP Test Report
IPQ and RDP Test Report
IPQ and RDP Test Report
IPQ and RDP Test Report
IPQ and RDP Test Report
IPQ and RDP Test Report
IPQ and RDP Test Report

More Related Content

PPT
powerpoint
PDF
Mini proj ii sdn video communication
PPT
Providing Controlled Quality Assurance in Video Streaming ...
PPT
A New QoS Renegotiation Mechanism for Multimedia Applications
PPTX
Application Layer Throughput Control For Video Streaming over HTTP2
PPTX
Dynamic Adaptive Point Cloud Streaming
PDF
Network
PPTX
HTTP Adaptive Streaming State of the Art and Challenges Ahead
powerpoint
Mini proj ii sdn video communication
Providing Controlled Quality Assurance in Video Streaming ...
A New QoS Renegotiation Mechanism for Multimedia Applications
Application Layer Throughput Control For Video Streaming over HTTP2
Dynamic Adaptive Point Cloud Streaming
Network
HTTP Adaptive Streaming State of the Art and Challenges Ahead

What's hot (13)

PPTX
Key Performance Indicators (KPI)
PPT
Nokia gsm-kpi-analysis-based-on-daily-monitoring-basis-presentation
PDF
Ericsson important optimization parameters
PPT
Keys to Enterprise WLAN Reliability
PPT
63077585 idle-mode-parameter-optimization
PPT
Ericsson optimization opti
PDF
Monitoramento transmissao hacje2014 03-07 00-50-27-_8.1
DOC
PDF
DPI-BASED CONGESTION CONTROL METHOD FOR SERVERS AND NETWORK LINES
PDF
On the Impact of Viewing Distance on Perceived Video Quality
PPTX
Tele2 - Companion App for_QOS_Telecom_Forum_eng 2015
PPT
House - Dynamic Bandwidth Throttling in a Client Server ...
PDF
7signal PoC AvansUoAS
Key Performance Indicators (KPI)
Nokia gsm-kpi-analysis-based-on-daily-monitoring-basis-presentation
Ericsson important optimization parameters
Keys to Enterprise WLAN Reliability
63077585 idle-mode-parameter-optimization
Ericsson optimization opti
Monitoramento transmissao hacje2014 03-07 00-50-27-_8.1
DPI-BASED CONGESTION CONTROL METHOD FOR SERVERS AND NETWORK LINES
On the Impact of Viewing Distance on Perceived Video Quality
Tele2 - Companion App for_QOS_Telecom_Forum_eng 2015
House - Dynamic Bandwidth Throttling in a Client Server ...
7signal PoC AvansUoAS
Ad

Similar to IPQ and RDP Test Report (20)

DOCX
TCP RemoteFX and IPQ
PPT
Network Application Performance
PPTX
chpater 4 FOR Information techonogy students
PDF
C08 – Updated planning and commissioning guidelines for Profinet - Xaver Sch...
PDF
Internet Path Selection on Video QoE Analysis and Improvements
PDF
A Review on Congestion Control Approaches for Real-Time Streaming Application...
PDF
Congestion control in packet switched wide area networks using a feedback model
PDF
Getting Started with callstats.io
PPT
Nss Labs Dpi Intro V3
PDF
Unlock your power with 4 dimensional actionable intelligence
PPTX
WANO - IND - Product Presentation
PPSX
Bluecoat Services
PPTX
Online TCP-IP Networking Assignment Help
PDF
5 maximazing networkcapacity_v4-jorge_alvarado
PDF
IR-Optimising-Your-Network.pdf
PDF
File_Transfer_Protocol_Design
PPT
Video Traffic Management
PPT
PV Powerpoint
PPTX
Performance Testing using LoadRunner
PPTX
Impact of Satellite Networks on Transport Layer Protocols
TCP RemoteFX and IPQ
Network Application Performance
chpater 4 FOR Information techonogy students
C08 – Updated planning and commissioning guidelines for Profinet - Xaver Sch...
Internet Path Selection on Video QoE Analysis and Improvements
A Review on Congestion Control Approaches for Real-Time Streaming Application...
Congestion control in packet switched wide area networks using a feedback model
Getting Started with callstats.io
Nss Labs Dpi Intro V3
Unlock your power with 4 dimensional actionable intelligence
WANO - IND - Product Presentation
Bluecoat Services
Online TCP-IP Networking Assignment Help
5 maximazing networkcapacity_v4-jorge_alvarado
IR-Optimising-Your-Network.pdf
File_Transfer_Protocol_Design
Video Traffic Management
PV Powerpoint
Performance Testing using LoadRunner
Impact of Satellite Networks on Transport Layer Protocols
Ad

Recently uploaded (20)

PDF
A novel scalable deep ensemble learning framework for big data classification...
PDF
Univ-Connecticut-ChatGPT-Presentaion.pdf
PPTX
cloud_computing_Infrastucture_as_cloud_p
PDF
project resource management chapter-09.pdf
PDF
NewMind AI Weekly Chronicles - August'25-Week II
PDF
1 - Historical Antecedents, Social Consideration.pdf
PPTX
KOM of Painting work and Equipment Insulation REV00 update 25-dec.pptx
PPTX
Digital-Transformation-Roadmap-for-Companies.pptx
PDF
Heart disease approach using modified random forest and particle swarm optimi...
PPTX
SOPHOS-XG Firewall Administrator PPT.pptx
PDF
DP Operators-handbook-extract for the Mautical Institute
PDF
Building Integrated photovoltaic BIPV_UPV.pdf
PDF
Profit Center Accounting in SAP S/4HANA, S4F28 Col11
PDF
7 ChatGPT Prompts to Help You Define Your Ideal Customer Profile.pdf
PDF
Web App vs Mobile App What Should You Build First.pdf
PDF
Getting Started with Data Integration: FME Form 101
PDF
Mushroom cultivation and it's methods.pdf
PDF
Enhancing emotion recognition model for a student engagement use case through...
PDF
Encapsulation theory and applications.pdf
PDF
Unlocking AI with Model Context Protocol (MCP)
A novel scalable deep ensemble learning framework for big data classification...
Univ-Connecticut-ChatGPT-Presentaion.pdf
cloud_computing_Infrastucture_as_cloud_p
project resource management chapter-09.pdf
NewMind AI Weekly Chronicles - August'25-Week II
1 - Historical Antecedents, Social Consideration.pdf
KOM of Painting work and Equipment Insulation REV00 update 25-dec.pptx
Digital-Transformation-Roadmap-for-Companies.pptx
Heart disease approach using modified random forest and particle swarm optimi...
SOPHOS-XG Firewall Administrator PPT.pptx
DP Operators-handbook-extract for the Mautical Institute
Building Integrated photovoltaic BIPV_UPV.pdf
Profit Center Accounting in SAP S/4HANA, S4F28 Col11
7 ChatGPT Prompts to Help You Define Your Ideal Customer Profile.pdf
Web App vs Mobile App What Should You Build First.pdf
Getting Started with Data Integration: FME Form 101
Mushroom cultivation and it's methods.pdf
Enhancing emotion recognition model for a student engagement use case through...
Encapsulation theory and applications.pdf
Unlocking AI with Model Context Protocol (MCP)

IPQ and RDP Test Report

  • 1. RDP 7.0 + IPQ TEST REPORTPrepared by IPeak Networks Incorporated22 October 2010<br />Executive Summary<br />Test Data Highlights – IPQ Improvements<br />Interpretation<br />WAN Scenarios<br />Satellite Scenarios<br />Retransmission Analyses<br />Conclusions<br />Full Test Results<br />IPQ is a very effective RDP optimizer for lossy networks<br />When measured in terms of bandwidth utilization IPQ improved RDP 7.0 performance in all tests where packet loss is greater than zero.<br />High bandwidth applications benefit greatly with up to 235% improvement in data transfer speeds to optimize application performance under lossy network conditions and provide corresponding improvements in the user experience, i.e., far fewer delays and a more fluid and productive experience.<br />Packet Loss and High Bandwidth Applications (e.g., PPT, Flash) <br />Loss and latency cause TCP “clamping” and reduced bandwidth. <br />Reduced bandwidth is experienced as reduced application performance.<br />IPQ- protects against loss<br />- prevents TCP clamping<br />- increases bandwidth, and<br />- restores application performance <br />Low bandwidth applications also benefit with the near complete elimination of TCP retransmissions and a significant reduction in the user’s experience of lag.<br />Packet Loss and Low Bandwidth Applications (e.g., Word) <br />Loss and latency triggers TCP retransmissions<br />TCP retransmissions are experienced as delayed application response or “lag”<br />IPQ- protects against loss<br />- reduces retransmissions, and<br />- reduces lag<br />IPQ IMPROVEMENTS - All Test Scenarios<br />WAN SCENARIOS<br />WAN and PowerPoint - Significant Reduction in Test Time, Increase in Average Speed<br />In these tests, packet loss caused significant delays, uneven application performance, and an unsatisfactory user experience. With IPQ protection against that packet loss, the delays were all but eliminated, the application performance was smooth and steady, and the user experience was comparable to the experience over the LAN.<br />Test Times Reduced<br />Over the WAN with 3% loss, this test required an additional 00:08:30 to complete when compared to the Calibration baseline test time of 00:41:01 [See Full Test Results]. With IPQ protection against that packet loss, 00:03:39 of that time was recovered.<br />The IPQ Improvement when the WAN suffers 5% packet loss is even more dramatic. At 5% packet loss, this test required an additional 00:27:36 to complete when compared to the baseline. With IPQ protection against packet loss, 00:21: 24 of that additional time was recovered.<br />With zero packet loss over the WAN, this test ran in 00:44:44. With 5% packet loss over the WAN, this test ran in 01:08:37. With IPQ protection against packet loss, this test ran in 00:47:13. With IPQ, the impact of packet loss on the test completion time was practically eliminated.<br />Speed (bps) Increased<br />IPQ improved average data transfer speeds in these tests. With IPQ Protection against 3% packet loss, average speeds were improved by more than 22%. At 5% packet loss, IPQ protection enabled an average data transfer speed improvement of more than 67%.<br />WAN SCENARIOS (con’t)<br />WAN and Word – Notable Reduction in Lag<br />Test Times, Speeds Unchanged<br />The measurements of test times, throughput, and average data transfer speeds in this test scenario confirmed that low bandwidth applications such as Word are not impacted by the bandwidth constraints induced by the interaction of packet loss and TCP. Test results with/without packet loss and with/without IPQ were not materially different.<br />Retransmissions Reduced, User Experience Enhanced<br />However, packet loss did cause delays in the application response times resulting in a notable deterioration in the user experience. With IPQ protection, the number of TCP retransmissions required to complete deliver user input to the application and return application response to the user was greatly reduced. I.e., the “lag” experienced by the user was significantly reduced. See the separate Retransmission Analyses for more. <br />WAN SCENARIOS (con’t)<br />WAN and Flash – Playout Times Corrected, Data Transfer Speeds Improved<br />Like PowerPoint, Flash is a high bandwidth application and as with the WAN PPTx tests packet loss caused significant delays, uneven and even un-useable application performance, and an unacceptable user experience in this test scenario. With IPQ protection against the packet loss, delays were greatly reduced, the application performance was smooth and steady, and the user experience was comparable to the experience over the LAN i.e., Flash was useable.<br />Test Times Reduced<br />Over the WAN with 5% loss, this test ran in 00:14:34, almost 6X the actual 00:02:30 runtime on the LAN [See Full Test Results]. With IPQ protection against the packet loss, this test time was reduced by more than half.<br />Speed (bps) Increased <br />IPQ improved average data transfer speeds in these tests. With IPQ Protection against 3% packet loss, average speeds were improved by more than 10%. At 5% packet loss, IPQ protection enabled an average data transfer speed improvement of more than 61%.<br />SATELLITE CONNECTION SCENARIOS<br />Satellite Connection and PowerPoint – Extreme Reduction in Test Time, Extreme Increase in Average Speed<br />As with the WAN PPTx tests, packet loss caused significant delays, uneven application performance, and an unsatisfactory user experience in the Satellite PPTx tests. With IPQ protection against that packet loss, the delays were reduced dramatically, the application performance was far more fluid steady, and the user experience was much closer to the experience over the LAN.<br />Test Times Reduced<br />Over a Satellite Connection with 3% loss, this test required an additional 01:07:36 to complete when compared to the Calibration baseline test time of 00:41:01 [See Full Test Results]. With IPQ protection against that packet loss, 00:52:46 of that time was recovered.<br />The IPQ Improvement when the satellite Connection suffers 5% packet loss is even more dramatic. At 5% packet loss, this test required an additional 01:25:03 to complete when compared to the baseline. With IPQ protection against packet loss, 01:08:18 of that additional time was recovered.<br />With zero packet loss over the Satellite Connection, this test ran in 00:55:27. With 5% packet loss over the Satellite Connection, this test ran in 02:06:04. With IPQ protection against packet loss, this test ran in 01:08:18. With IPQ, the impact of packet loss on the test completion time was dramatically reduced.<br />Speed (bps) Increased<br />IPQ improved average data transfer speeds in these tests by as much as 2.3X. With IPQ Protection against 3% packet loss, average speeds were improved by more than 97%. At 5% packet loss, IPQ protection enabled an average data transfer speed improvement of more than 234%.<br />SATELLITE CONNECTION SCENARIOS (con’t)<br />Satellite Connection and Word – Notable Reduction in Lag<br />Test Times, Speeds Unchanged<br />The measurements of test times, throughput, and average data transfer speeds in this scenario confirm that low bandwidth applications such as Word are not impacted by the bandwidth constraints induced by the interaction of packet loss and TCP. Test results with/without packet loss and with/without IPQ were not materially different.<br />Retransmissions Reduced, User Experience Enhanced<br />However, packet loss did cause delays in the application response times resulting in a notable deterioration in the user experience. With IPQ protection, the number of TCP retransmissions required to deliver user input to the application and return application response to the user was greatly reduced. I.e., the “lag” experienced by the user was significantly reduced. See the separate Retransmission Analyses for more.<br />SATELLITE CONNECTION SCENARIOS (con’t)<br />Satellite Connection and Flash – Playout Times Improved, Data Transfer Speeds Doubled<br />Like PowerPoint, Flash is a high bandwidth application and as with the Satellite Connection PPTx tests packet loss caused frequent, prolonged delays, and a largely un-useable user experience in this test scenario. With IPQ protection against that packet loss, the delays were greatly reduced, and the application performance was improved as was the user experience.<br />Test Times Reduced<br />Over the Satellite Connection with 5% loss, the Flash test completed in 00:37:26, many times the actual Flash running time of 00:02:30 on the LAN [See Full Test Results]. With IPQ protection against packet loss, the test time was reduced by a full 00:22:16 bringing Flash over a Satellite Connection into the realm of possibility.<br />Speed (bps) Increased<br />IPQ improved average data transfer speeds in these tests. With IPQ Protection against 3% packet loss, average speeds were improved by more than 121%. At 5% packet loss, IPQ protection enabled an average data transfer speed improvement of more than 118%.<br />RETRANSMISSION ANALYSES<br />WAN and Word – Dramatic Reductions in Retransmissions and Lost Keystrokes<br />The analysis of the retransmissions with and without IPQ was prompted by the observation of high levels of lag or delays in the application response to user input, i.e. to the typing and scrolling activities. Keyboard input that fails to reach the virtual machine and application due to packet loss requires retransmission. The application response is delayed until the retransmission is made successfully and the keystroke is communicated to the application. Similarly, the application response to the keystroke is also subject to potential loss, also requiring retransmission from application to user. Whether lost enroute to or from the application or the user – or both – the loss and retransmission manifest as lag.<br />TCP retransmissions are at once symptomatic of packet loss and the programmatic response to packet loss. IPeak Networks undertook this additional retransmission analysis to provide a set of underlying measurements that may be correlated to the user experience of lag.<br />Retransmissions and Lost Keystrokes Reduced<br />At 3% packet loss, while typing and scrolling over the WAN, retransmissions accounted for more than 11.47% of the data transferred. In effect, approximately one of every nine keystrokes was lost due to packet loss and required retransmission. With IPQ protection against the packet loss, the retransmissions accounted for 0.57% of the data transferred, a net reduction in retransmissions equal to more 95%. With IPQ, the number of keystrokes lost due to packet loss was reduced to one in 200.<br />When typing and scrolling at 5% packet loss, retransmissions accounted for 14.89% of the data transferred. IPQ protection reduced the retransmissions in this test by almost 90% and the lost keystrokes were reduced from about one in six to fewer than one in sixty.<br />RETRANSMISSION ANALYSES (con’t)<br />Satellite Connection and Word – Dramatic Reductions in Retransmissions and Lost Keystrokes<br />The analysis of the retransmissions with and without IPQ was prompted by the observation of high levels of lag or delays in the application response to user input, i.e. to the typing and scrolling activities. Keyboard input that fails to reach the virtual machine and application due to packet loss requires retransmission. The application response is delayed until the retransmission is made successfully and the keystroke is communicated to the application. Similarly, the application response to the keystroke is also subject to potential loss, also requiring retransmission from application to user. Whether lost enroute to or from the application or the user – or both – the loss and retransmission manifest as lag.<br />TCP retransmissions are at once symptomatic of packet loss and the programmatic response to packet loss. IPeak Networks undertook this additional retransmission analysis to provide a set of underlying measurements that may be correlated to the user experience of lag.<br />Retransmissions and Lost Keystrokes Reduced<br />Retransmissions represented 11.49% of the data transferred when typing and scrolling over a Satellite Connection with 3% packet loss. One in nine keystrokes was lost. With IPQ protection against packet loss, and retransmissions were reduced by almost 90 percent and lost keystrokes were reduced by from one in nine to one in 62. At 5% packet loss, retransmissions were reduced by more than 80% and lost keystrokes were reduced from approximately one in 6 to fewer than one in forty.<br />CONCLUSIONS<br />The results of these tests confirm that packet loss and latency, both individually and in combination, have a seriously damaging effect on the performance of both high bandwidth and low bandwidth applications running in virtual desktops. It follows that it is desirable to make efforts to reduce latency and reduce packet loss in order to preserve applications performance and support a satisfactory user experience of the virtual desktop.<br />There is little that can be done to reduce network latency other than locating the data centers and remote desktop users in close physical proximity to one another, a strategy that is antithetical to the vision of a ubiquitous and even global deployment of virtualized computing. It is therefore reasonable to focus on the performance gains and user experience benefits that may be realized through the use technologies that reduce packet loss.<br />The results of these tests further confirm that IPQ is an effective solution for the problem of network packet loss. When measured in terms of reduced test time-to-completion, increased thoughput, increased average data transfer speeds, and reduced TCP retransmissions, IPQ proved to be very effective in optimizing RDP 7.0 performance in all tests where packet loss is greater than zero.<br />High bandwidth applications benefited greatly with up to 235% improvement in data transfer speeds helping to optimize application performance under lossy network conditions and provide corresponding improvements in the user experience, i.e., far fewer delays and a more fluid and productive experience.<br />Low bandwidth applications also benefit with the near elimination of TCP retransmissions, a vast reduction in the number of keystrokes lost as a function of packet loss, and a corresponding reduction in the user’s experience of lag.<br />-8204201109980<br />