Abstract: Because there is a delay between fault detection and fault removal in open source software, current researchers generally establish the corresponding open source software reliability model by investigating the fault detection process. However, there will be significant errors in estimating software reliability and the number of remaining failures. Especially, it is inaccurate to estimate the parameter value of the model by the number of detected faults. Because there are many problems in the detected faults, such as repeated faults, failure that cannot be repeated, failure that cannot be removed and false failures, etc. Considering this problem, this paper proposes to establish a corresponding open source software reliability model by considering the troubleshooting process. Experimental results show that the proposed model is effective and more accurate in predicting the number of residual faults. The proposed model can help developers to evaluate the reliability of open source software and predict the number of remaining faults in open source software.