How would you handle a situation where a reported issue cannot be reproduced in your test environment?
