You are here: Manual / Special / Remote


Using remote SEAMCAT server to perform simulation - DISCONNECTED WITH SEAMCAT 4 (MAY 2012)

Note: the remote server has been disconnected in May 2012 until further notice.

Note that you can set-up your own machine (or a machine within your organisation) as a remote server.

Remote computing is a function implemented in SEAMCAT-3, which together with Batch function ease the performing of lengthy simulations, such as simulations with many interfering links or with many snapshots involving CDMA systems.

The idea with remote computing is that after creating the scenario, the user may send it to a dedicated server that will carry out lengthy simulations, and afterwards would download the results at any convenient time. The server may be any PC connected to a network, which is having a copy of SEAMCAT-3 software installed and launched in Server mode. Users may setup such servers themselves (e.g. on a company network), or to use central server set-up by ECO and accessible from anywhere over the Internet.

The flow-chart illustrating the process of remote computing is shown below:

Both individual workspaces and a batch list can be sent to the remote server. Although the process is similar it is started in two different ways.

  • For a single workspace the process of sending it to remote server is initialised throug the Tools menu by selecting Remote Computing / Send current Workspace to server.
  • For batch the process is started by clicking Send batch to server at the bottom of Batch management window.

Either way, the following window will appear:

The default values of this dialog window specify the server address and port of the SEAMCAT calculation server hosted by ECO. The ECO server does NOT require a password, so the user will have to enter an e-mail address where notifications could be sent by the server. When clicking ok, the job will be sent to the server and the user may close SEAMCAT program altogether and turn to other jobs. When the remote server has finished processing the job, an e-mail will be sent to the specified e-mail address.

Note: If the user wants to use a different computation server (e.g. an own server set up by the user), then the appropriate address/port for that different server must be specified in the above window, including the password if relevant.

To retrieve the calculation results from the server the user proceeds via command Remote Computing / Check status on remote job from the Tools menu. This will open a dialogue with a list of all jobs that have been sent by the user to the remote servers.

By clicking Update status, the user can ask the server for a status of the job. When the server reports that the job has been completed the Update status button will change to a Download results button. Clicking this will start download process. Depending on the user's scenario, the results can contain quite a lot of data, so the download time may vary depending on the scenario and the internet connection speed. Upon succesfull download, the programme will ask if the user wants to open the workspace with results. In case of remote calculation of batch jobs, their calculation results may be downloaded from the Batch management window.

Note: The remote computing protocol is designed so that only the dRSS, iRSS (as well as the cell loading vectors for scenarios with CDMA victim) are stored on the server and sent back to the user. Therefore no graphical information (such as EGE plots, CDMA system plots) will be available for analysis by the user.

Partly because of this, and partly because of a general lack of user control of simulations sent to the server, it is strongly recommended that the user performs a trial EGE run (e.g. 10 events) of the created scenario on the local PC before sending it to the server. This would allow the user to check the stability of EGE simulations (the time required per snapshot, no exceptions thrown, no endless loops created, any plug-in properly called and so on), to ascertain mutual positioning of modelled system, and to check the tendency of produced results.

Note: It is also possible to send remotly a scenario, which contains calling of Plug-ins (either Propagation model plug-in or Post-processing plug-in). However this requires that the relevant plug-in had been pre-installed on the remote server! . Therefore, if you want to use the remote computation with your plug-ins on the central ECO server, please contact Please contact seamcat @ to inquire about possibility of uploading your plug-ins to the server.