PENINGKATAN AVAILABILITY INFRASTRUKTUR STORAGE PADACLUSTER COMPUTING MENGGUNAKAN METODE FAULTTOLERANCE
Storage type in cluster computing infrastructure consists of object-level storage, file-level storage, and block-level storage. Storage infrastructure network based on cluster computing divided into Storage Attached Network (SAN) and Network Attached Storage (NAS). A storage cluster system is inseparable from fault or system failure. Failure can not be avoided but reduced the frequency and robustness of the system had to facilitates a strong and able to adapt to such failure. In the aspect of security triad known as confidentiality, integrity and availability that the both of them are related with this research. This output of the research should be able to suppress Mean Time To Repair (MTTR) value and increase Mean Time Between Failure (MTBF) value to get maximum availability. Fault tolerance offers redundancy as a robust enhancement solution for system using replication and erasure coding. In addition, the researchers will also calculate the recovery time required system or software, or existing protocol against block, file, and storage objects to get the smallest MTTR value. In addition to considering the availability of a system, which needs to be considered is throughput, capacity, computer resources and cost effectiveness. The experiments conducted by the authors are represented in the form of models so they can be used for cluster use in general and close to the original implementation.