To maintain consistency on a distributed database system, a commit protocol is often employed to ensure that either all the subtransactions of a transaction commit or all abort. Special care needs to be taken to ensure the correct functioning of a commit protocol in the face of node failures during the execution of the protocol. A protocol is said to be fault-tolerant if it properly commits or aborts, even if nodes fail or network partitions. In this paper, we present three fault-tolerant decentralized commit protocols which are the extensions of the optimal decentralized commit protocol (ODCP) presented in [141. The first protocol (FTDCPI) ensures that all sites can achieve a consistent final decision by at most four extra messages in the single site failure situation. The second protocol (FTDCP2) deals with multiple site failures. The third protocol (NFTDCP), a simplified version of the FTDCP2, is resilient to some types of network partitionings.