An alternative to centralized deadlock detection is the building of a hierarchy of deadlock detection detectors. Deadlocks that are local to a single site would be detected at that site using the local WFG. Each site also sends its local WFG to the deadlock detector at the next level. Thus, distributed deadlocks involving two or more sites would be detected by a deadlock detector in the lowest level that has control over these sites. For example a deadlock at site 1 would be detected by the local deadlock detector at site 1and 2 then DD11 detect it. Finally if the deadlock involves sites 1 and 4, DD00 detects it where x is either one of 1,2,3 or 4.
The hierarchical deadlock detection method reduces the dependence on the central sites, thus reducing the communication cost. It is however, considerably more complicated to implement and would involve nontrivial modifications to the lock and transactions manager algorithms.
2) Distributed deadlock detection algorithms delegate the responsibility of detecting deadlocks to individual sites. Thus as in the hierarchical deadlock detection there are local deadlock detectors at each site which communicate their local WFGs with one another. Above all R* seems to be the more widely and referenced.
The hierarchical deadlock detection method reduces the dependence on the central sites, thus reducing the communication cost. It is however, considerably more complicated to implement and would involve nontrivial modifications to the lock and transactions manager algorithms.
2) Distributed deadlock detection algorithms delegate the responsibility of detecting deadlocks to individual sites. Thus as in the hierarchical deadlock detection there are local deadlock detectors at each site which communicate their local WFGs with one another. Above all R* seems to be the more widely and referenced.
