Export (0) Print
Expand All

Performance Considerations

Microsoft Dynamics Nav 2009

You can use the information in this section to help decide whether the RoleTailored client over WAN option is appropriate for your organization. Specifically, this section should help you answer the following questions:

  • Do your users’ usage scenarios support running the RoleTailored client over a WAN?

  • What kind of performance can you expect when you run the RoleTailored client over a WAN?

Targeted Usage Scenarios

There are two primary scenarios for running the RoleTailored client over a WAN.

  • Mobile light-usage users. Salespeople who visit customers and look up order histories and take orders when they are at the customer site are examples of mobile light-usage users. These users want quick and easy access to data to support their primary tasks. This usage is typically not extensive, and users prefer to not have to log on to the company network through a virtual private network.

  • Hosted users. Multiple users who sit at one location and access Microsoft Dynamics NAV Server at a remote location are hosted users. These users do not have control of the installation, which is managed by a hosting provider.

For each scenario, the following assumptions are made:

  • Each user has a download speed of at least 1 megabit per second and an upload speed of at least 200 kilobits per second.

  • Network latency is less than 200 milliseconds.

  • The RoleTailored client is running on a computer that cannot log on to an Active Directory domain controller, either in the domain where the RoleTailored client is installed or in the domain where Microsoft Dynamics NAV Server is installed. The RoleTailored client and Microsoft Dynamics NAV Server can be in the same Active Directory domain or in separate domains.

Performance Testing Results

Performance testing of these scenarios focused on latency and bandwidth. An environment where both latency and bandwidth could be throttled helped simulate different types of connectivity. The tests were based on 10 concurrent users posting 10 automated one-line sales orders with the RoleTailored client. This kind of measurement is not completely realistic because the order entry is so fast that the screen cannot render before the Post button is pressed. Therefore, if this is the benchmark load, then any real-life loads that involve similar operations will be slower.

Bandwidth that ranged from a 10-megabit download and 1-megabit upload connection to a 300-megabit download and 300-megabit upload connection, which is comparable to a local area network, was tested.

Latencies that ranged from 0 milliseconds to 600 milliseconds were also tested. This range approximates connections from a fast LAN connection to a slow ADSL connection. The range also approximates a fast satellite connection, which would be between 500 milliseconds and 1000 milliseconds.

In the following graph, the x-axis from left to right shows the round-trip time added to the connection in milliseconds. The line that runs from the lower-left to the upper-right shows the latency, or response times. The other lines show the maximum kilobytes received per second, average kilobytes received per second, maximum kilobytes sent per second, and the average kilobytes sent per second.

Throughput and latency for RTC over WAN

The graph shows that latency linearly affects the response time. It also shows that a higher latency affects the ability to use the available bandwidth and that the "elbow" or optimum latency is between a latency of 100 milliseconds and 150 milliseconds.

In the following graph, the x-axis shows bandwidth per user, and the y-axis shows the response time for the 10 sales orders. On the bandwidth per user axis, 5/1 represents a 50-megabit upload and 10-megabit download connection, and 2/0,5 represents a 20-megabit download and 0.5 megabit upload connection.

Bandwidth against response time

The graph shows that the "elbow" or optimum bandwidth is between 1.50/0.3 megabits and 2/0.1. Additional tests show that the determining factor for these connections is the upload speed rather than download speed, and that the elbow is between 0.1 and 0.3 megabits per user for the tested scenarios.

Because these tests targeted limited bandwidth scenarios, you can also have a single session use all available bandwidth for any kind of connection when you transfer a large file or run a large report.

For more information, see the "Microsoft Dynamics NAV 2009 R2 Hot Topic: RoleTailored Client for Remote and Roaming Users" session on the Partner Learning Center (requires Microsoft Partner Network login).

Community Additions

ADD
Show:
© 2014 Microsoft