The error can occur when logging onto TeamInspector when it is attempting to retrieve the projects list.
This can happen when the number of active streams are more than can be handled within the SCM interval monitor time and thus creates a backlog.
Best practise would be to increase the "SCM monitor poll threads" value in the TeamInspector configuration, this affects the number of worker threads that check on the status of the streams. Also increase the "SCM interval monitor" value, as this will give the threads more time to process the stream queue before timing out.
For example, if the interval is set to 30 seconds and there are a large number of streams, this may create problems. Setting the interval between 300 - 600 seconds will result in the stream being validated for changes every 5-10 minutes. Increasing the thread count alone will create a lot of StarTeam activity which is why a combination of both options is recommended.