<![CDATA[Microservice architecture as a solution to the monolithic architectural system is growing more and more in use today. As a fairly successful solution, the microservice system is not devoid of new problems and other challenges that arise. Microservice as a self-developed system and a service that communicates between networks are often fault. Any fault that occurs can be caused by high latency or because of the internal problem that causes the slowing down process. This may cause microservice to be inaccessible to the user or other microservice in need. The research provides a solution to the implementation of a microservice system that can withstand fault by applying the stability pattern with a circuit breaker and bulkhead pattern. Tests are made on the implementation of microservice systems using the PRS case study used by the Christian university of Petra. Fault executives are implemented to test the comparative performance of a DFTM method of circuit breaker with a circuit breaker, a bulkhead with no bulkhead, and to measure the fault levels microservice by MRMM. Based on the results of the tests, DFTM circuit breaker performance was at 46% faster than normal. The results of the testing also have the subtractable use of bulkhead could prevent resource use more than the capacity of microservice servers had if the latency was higher. Testing result also mentioned that the fault tolerance of the microservice measured by MRMM violating resilience of the benchmark. This can be proven from declining success rates, increased average response time, and decreased transactions per second.]]>
