Книга: Distributed operating systems


Page-based DSM takes a normal linear address space and allows the pages to migrate dynamically over the network on demand. Processes can access all of memory using normal read and write instructions and are not aware of when page faults or network transfers occur. Accesses to remote data are detected and protected by the MMU.

A more structured approach is to share only certain variables and data structures that are needed by more than one process. In this way, the problem changes from how to do paging over the network to how to maintain a potentially replicated, distributed data base consisting of the shared variables. Different techniques are applicable here, and these often lead to major performance improvements.

The first question that must be addressed is whether or not shared variables will be replicated, and if so, whether fully or partially. If they are replicated, there is more potential for using an update algorithm rather than on a page-based DSM system, provided that writes to individual shared variables can be isolated.

Using shared variables that are individually managed also provides considerable opportunity to eliminate false sharing. If it is possible to update one variable without affecting other variables, then the physical layout of the variables on the pages is less important. Two of the most interesting examples of such systems are Munin and Midway, which are described below.

Оглавление книги

Оглавление статьи/книги

Генерация: 0.033. Запросов К БД/Cache: 0 / 0
Вверх Вниз