It doesn't sound like your managment server is processing messages properly. The queue should be relatively small.
Are you getting any new messages in your activity monitor?
It is going to take a long time to process 2GB of message in your queue. You can eathier reboot and wait while your server processes the messages (you will notice time created and time recieved will be different, as there is a delay in processing) or you can stop the services delete the queue file, and then start the services back up. The queue file will be re-created and all the old messages in the queue will be lost.
Take a copy of your Q folder ... use the HP binary to read the Q file. It could be because of a message storm from single node by opcle/opctrapi or subnet out of network for long time ...... delete the qfiles and let the mgmt server create new one upon restart of om processes