An issue that the real-time software research community has long recognized as an important technological challenge but has not shown much progress in meeting the challenge is to guarantee response times of real-time distributed systems. Two basic problems must be addressed to effectively meet this challenge: establishment of the distributed real-time program structure and the system infrastructure structure that enable systematic analysis of the worst-case time behavior of the application systems; and development of tools which perform automated analysis of the worst-case time behavior while leaving only minimal work to the designers. Research on the first problems has produced some useful results in recent years. Time is thus ripe for tackling the second problem on the basis of those recent developments in handling the first problem. This paper describes a desirable systematic timing analysis process which builds upon the recent results in the first problem area. An outline of a desirable analysis tool is also presented.