Increasing Processing Speed

Idle Sensitivity -- This is an often overlooked setting that can be adjusted in Windows 9x to increase the speed of work unit processing. In simple terms, "Idle Sensitivity" controls how sensitive Genome@home is to requests for CPU time from other applications, processes and services. By default, Genome@home is installed with a medium idle sensitivity setting. By lowering idle sensitivity (moving the slide bar to the left), Genome@home will run at a higher priority more frequently, thus processing work units more quickly. Most users won't notice an overall performance difference by running Genome@home at a lower idle sensitivity. To access this setting, go to c:\windows\system and find the file named conagent.exe (you will also see there an ms-dos icon called conagent, but you want the file conagent.exe). Right click on that file, choose "Properties", then "Miscellaneous". You will see the idle sensitivity setting in the column on the left. Note that changing this setting will affect all DOS programs that run on your computer. If that is an issue and you would like to adjust idle sensitivity for Genome@home only, copy the conagent.exe file to your Genome@home directory and make the changes to that file. It is then mandatory that you click on the MS-DOS icon at the top of the Genome@home program window, select "Properties", then the "Program" tab. Edit the Command line entry to point to the path of the conagent.exe file that was copied to your Genome@home directory.

Another way to accomplish this is to use a 3rd party utility called "TaskInfo2000 ". More information is available about this utility, including a download link, in the Software Utilities section of this web site.

Making Life Easier for Modem Users

-- Version 0.98 of the Genome@home software has added some functionality that makes it easier and more efficient for modem users to work with the software. In the Genome@home menu, there is now a "Run Off Network" option. The efficiency gained by this option is that if a work unit has been completed, the software will not delay the reprocessing of that work unit while it intermittently checks for an internet connection. Of course, if a work unit is reprocessed, it will produce a viable result that counts as an additional work unit. So, of course, eliminating the down time while the software checks for an internet connection will increase work unit production.

-- Version 0.98 also adds a menu option called "Upload Results". Here's how to use that option. When you are ready to upload a result, close the Genome@home software gracefully by using control-c. Then, while you are connected to the Internet, select "Upload Results" from the program menu. The software will open a DOS window so that you can see the upload, and will close upon completion. If you then reopen the Genome@home client, while still connected to the internet, a new work unit will be downloaded and processing will continue. At that point, you can close your internet connection. Note: if you missed the completion of a work unit, it is not necessary to allow the unit to be completely reprocessed. If you close the client when it has completed a protein sequence, and follow the above procedure, the new work unit will be processed beginning with the last sequence you had completed before uploading your results. No data is lost.

-- If you would like the client to initiate an unattended dial-up connection upon completing a work unit, go to the Control Panel. Find the "Internet Options" icon, open it, and find the "Connections" tab. With your ISP highlighted, select "Dial whenever a network connection is not present". The same setting can be accessed in Internet Explorer by selecting "Tools", "Internet Options ", and then "Connections". To configure your system so that the modem will disconnect, go to Dial Up Networking, right click the icon that represents your ISP, select Properties, and enter the tab called "Dialing". You will see the check box called "Enable idle disconnect", and you can set the idle time after which the modem will disconnect. In this fashion, it is easy to configure you system to perform unattended uploads and downloads. Reminder: You must let Windows save your dial-up connection password for this approach to work. Otherwise, the dial-up procedure will hang at the password prompt.

"Sneakernetting" the Genome@home Program

Sneakernetting refers to running Genome@home on a computer that has no internet connection and then uploading those results from an internet-connected PC. The Genome@home program includes the option of processing work units (wu's) without a network connection (the "no-net" option). The issues become how to "feed" the Genome@home application its initial work unit (wu) for processing, and then how to upload the finished wu from a internet-connected PC.

Note that you can run Genome@home in the sneakernet mode on as many computers as you wish, but you must already have Genome@home installed on at least one computer that does have an internet connection ( the "networked" computer) so that you can upload the completed wu's from the no-net computers(s).

There are three steps involved in sneakernetting Genome@home:

1. Installing Genome@home on the non-internet-connected ("no-net") computer(s),
2. Copying a wu to the no-net machine so that it can begin processing, and
3. Following the steps used to upload the completed results from another, internet- connected PC.

Installing the Genome@home Program:

1. Download the program installation file from Stanford. Installation instructions can be found on the Stanford web site as well as on the KWSN-Genome web site under the Getting Started section.
2. Obtain a wu that is ready to process. Obtain the Zip file here. Unzip the file and copy all of its contents into your "C:\Program Files\Genome@home" folder (default location). Copy the file named ghclient.cfg (probably located in the "C:\Program Files\Genome@home" folder) from your "networked" computer to your no-net computer's Genome@home folder (the .cfg file identifies you as the person who has processed the wu). Or, to save you some time, if you're going to be sneakernetting several computers, copy the ghclient.cfg file into the Zip file above.
3. Run the Genome@home program in -nonet mode (Start/Programs/Genome@home/Run off network). You are now processing the wu. When the program is finished processing a wu, the Genome@home client will create three files (csum.###, output.chi.###, and str.###), all with the same numeric (###) extension. You can let the program run for as long as you wish and it will continue to produce completed, valid wu's.
4. When you are ready to upload the completed results, copy all of the .### files to the Genome@home directory on the computer that has an internet connection, and run the client with the -upload option (Start/Programs/Genome@home/Upload results). If Genome@home is already running on this computer, it will report that another client is running and ask if you wish to proceed. Type "yes" and hit "enter". Genome@home will upload all the completed wu's and then close itself while leaving the original client running.

Additional notes: If you are running sneakernet on multiple computers, and copying the results from those computers to the networked machine's Genome@home directory for subsequent uploading then, on occasion, you may get a warning that the .### extension on the files being copied (from the no-net to the net computer) match those of existing files. You will be asked if you wish to overwrite the original files. Just reply "no" and copy those "duplicate" files into a temporary directory. Upload the results you have already copied into the Genome@home directory on the networked computer, then copy those duplicate files from the temporary directory to the Genome@home directory and upload them separately (via the upload results option explained above). The duplicate files are duplicates in name-only. The results are different and will be counted in your production totals. The Genome@home program has a limited # of file names, so multiple computers running no-net, on occasion, produce files with the same file name.

Caching: If you can't connect to the Stanford server (it's down, or you're offline) when it's time to deposit finished work, you'll get a few PutWork error messages. After twenty minutes, Genome@home will stop attempting to connect, store the results on disk, and rerun the existing work units. Since the results are different each time Genome@home reprocesses the same wu, this is completely equivalent to downloading new work units. Genome@home will continue rerunning work units (indefinitely) until it is able to connect to the server, at which time it will upload all the results to the server, and you will get credit for the total of all the work units processed.

Running Genome@home Minimized

Does the Genome@home program window open on your screen at startup? If you would like the program to run in the taskbar after starting up, click on the MS-DOS icon at the top left of the program window. Select "Properties", then "Program". Change the "Run" option from "Normal Window" to "Minimized".

Want to Eliminate the "Close Filter" Warning at Restart?

Following the instructions from the above section, find the conagent.exe file, right click it and select "Properties" then "Miscellaneous". Uncheck the box that says "Warn if still active". Now, Windows will close Genome@home and restart normally. This setting affects all DOS programs that run on your computer. If you wish to make this change for Genome@home only, copy conagent.exe to your Genome@home directory, make the appropriate changes, then open the application and click on the MS-DOS icon in the top left corner. Select "Properties" and in the "Program" tab, edit the Command line to point to the path of the conagent.exe file that was copied to your Genome@home directory. Thanks to VIRUS62 for contributing this tip.