Sometimes, especially if you run server side analysis, the process can take several minutes: in that situation you may want to start the process, go back to your build and collect the results later. This can be done passing the
--start-only argument to the client. In that situation the client will collect the required data, send them to the server and return to the shell. You can then run your build, executing it in parallel to the work done on the Meterian servers. At the end of your build you just need to launch again the client with the same configuration, removing the
--start-only argument: it will pick up the same analysis again, contact the server and return the results. Of course, if the server is still not ready, you may need to wait until everything is ready, but the client will report about the progress.