Ok, then I guess we'll increase the timeout and cycle the client from time to time. Thanks for the input @yetitwo 🙂