Labs

Remote access software

The overall speed of remote access is difficult to judge due to the myriad of connections possible and their supported transfer rates. For example, remotely controlling a PC over a standard Ethernet network is going to be quicker than running the session over a 56K modem Internet link. The one task that's likely to take the longest and have the biggest impact on your patience is file transfer, so we opted to test this feature on each review product.

To achieve a best-case scenario, we limit the test to a connection between two systems on a dedicated LAN. For the guest system, we use a 1.13GHz Pentium III server equipped with 512MB of RAM and running Windows 2000 Server SP 3 with Internet Explorer 6 SP 1 installed. At the centre of the network is an Allied Telesyn AT-9410GB Gigabit Ethernet switch and the server is connected via a 1000BaseT network adaptor. The host system is a Pentium III/866 server with 768MB of RAM - also with Windows 2000 Server SP 3 installed - while the network connection is through a 10/100BaseTX network adaptor. Prior to each test, both servers are refreshed with a disk image of a clean OS installation. Although initial tests indicated there was little to be gained in terms of performance, we also switched off any encryption on each product.

The first test is a simple copy of a 571MB VideoCD format file from the guest to the host using each product's file-transfer tools. As the file is already compressed, any other techniques offered by each product will have no effect on performance. The second test involves copying a folder to the host containing a 602MB mixture of bitmaps, documents, spreadsheets, databases and PowerPoint presentations.

You'd think each remote access product would be able to copy a file from one PC to another in roughly the same amount of time, and it came as a surprise to see how much variation there was. Altiris' Carbon Copy Solution is marked as having failed the tests, as it took 50 minutes to copy the video file and 37 minutes to copy the mixture. As Expertcity's GoToMyPC only operates over the Internet through the company's website, we chose not to include any comparative test results as it would be unfair. However, out of curiosity we did try copying the video file and estimate that it would have taken about eight hours to transfer.

Performance is only one area we consider when choosing the best product. The ease with which the software can be deployed and used will be important, as it should be designed to help and not hinder support staff. The number of features also comes into the equation, as we find this varies considerably, particularly with remote-control tools, where some products merely offer a window to the host system and no options for drawing tools, sending keystrokes or rebooting.