Contorl And Management Of Dual Controller RAID System Base On Logic Volume

In current data storage, disk array has been widely used, the diskarray controller has been a single point of failure of the RAID(Redundant Array of Independent Disks) system. To improve the availability of storage system, we can use two controller as redundant to resolve such failures, in order to improve system sensitivity, the dual control of storage management is often manage simple configuration. On the other hand, human and intelligence management of massive data management is to be a big demand,the development of storage virtualization management is also increasingly becoming a hot spot.The current RAID management software on dual-controller,due to complexity management,system generally do not provide a virtual device level management,and host level virtualiation management also increased the complexity and difficulty,and affect the data service reliability and performance.We research and analysis mainly from the following aspects.First,ensure the consistency of data state on double-controller one by separating the metadata of virtual devcie and communication with configuration independently.Second,through the heartbeat monitoring and status notifications between the two controller, system ensure prompt response time after controller failure. Third, we design the rational self-fault detection and state transition to ensure that the controller will switch in a closed state. Fourth,we resolve the brain spilt for heartbeat failure by the disk control competition to and improve the system design recovery process by practical details.We implementation the management of dual-controller system based on logical volume, and achieved timely and effective monitoring, treatment and recovery, at the last the system was testing by fault simulation.