Semi-Preemptible Locks for a Distributed File System
نویسندگان
چکیده
Many applications require the ability to obtain exclusive access to data, where an application is granted privileges to an object that cannot be preempted and limits the actions of other processes. Local file systems support exclusive access to files by maintaining information about the access rights of current open file instances, and checking subsequent opens for compatibility. Implementing exclusive access in this manner for distributed file systems degrades performance by requiring every open file to be registered with a server that maintains global open state. We introduce a distributed lock for managing file access, called a semi-preemptible lock, that avoids this performance limitation. Clients locally grant open requests for files that are consistent with a distributed semipreemptible lock that they hold. File system clients retain or cache distributed locks, even in the absence of open file instances. When a file access lock is already cached, a client services open requests without a server message, improving performance by exploiting locality, the affinity of files to clients.
منابع مشابه
Design and Implementation of Semi-preemptible IO
Allowing higher-priority requests to preempt ongoing disk IOs is of particular benefit to delay-sensitive multimedia and real-time systems. In this paper we propose Semi-preemptible IO, which divides an IO request into small temporal units of disk commands to enable preemptible disk access. We present main design strategies to allow preemption of each component of a disk access—seek, rotation, ...
متن کاملHierarchical Replication Control
We present a hierarchical locking algorithm that dynamically elects a primary server in a replicated file system at various granularities. We introduce two lock types: shallow locks that control a single file or directory, and deep locks that lock everything in the subtree rooted at a directory. Experimental results show that for typical use cases, deep locks can make the overhead of replicatio...
متن کاملSafe Caching in a Distributed File System for Network Attached Storage
In a distributed file system built on network attached storage, client computers access data directly from shared storage, rather than submitting I/O requests through a server. Without a server marshaling access to data, if a computer fails or becomes isolated in a network partition while holding locks on cached data objects, those objects become inaccessible to other computers until a locking ...
متن کاملDistributed Lock Manager
In this paper, we describe Distributed Lock Manager to provide reader with thorough insight into the functionality and evolution of distributed lock manager. Apart from the fact that at present, different solutions for implementation of DLM are available in the market, the insight presented applies to all, in general. We also describe Chubby lock service which is intended to provide coarse-grai...
متن کاملImplementing MPI-IO Atomic Mode and Shared File Pointers Using MPI One-Sided Communication
The ROMIO implementation of the MPI-IO standard provides a portable infrastructure for use on top of a variety of underlying storage targets. These targets vary widely in their capabilities, and in some cases additional effort is needed within ROMIO to support all MPI-IO semantics. Two aspects of the interface that can be problematic to implement are MPI-IO atomic mode and the shared file point...
متن کامل