Electrical computers and digital data processing systems: input/ – Access locking
Reexamination Certificate
2000-11-28
2004-06-29
Portka, Gary (Department: 2188)
Electrical computers and digital data processing systems: input/
Access locking
C710S220000, C711S152000
Reexamination Certificate
active
06757769
ABSTRACT:
FIELD OF THE INVENTION
This invention relates generally to a method and apparatus for improving performance in systems where multiple processors contend for control of a shared resource through a lock associated with the shared resource, and more particularly to a method and apparatus for improving performance in intelligent data storage systems.
BACKGROUND OF THE INVENTION
When a computer system resource is shared by multiple processes running on multiple processors, or even on one processor, often there must be some way of insuring that no more than one such process may access that resource at any one time. In designing complex data storage systems including multiple processors, synchronizing access to shared resources has been recognized as an issue which must be addressed in order to maintain the consistency and validity of the data. However, the sharing issue may arise in connection with almost any resource that might be used by multiple requestors.
Many high-performance storage systems are intelligent data storage systems which may be accessible by multiple host computers. These may include, in addition to one or more storage device arrays, a number of intelligent controllers for controlling the various aspects of the data transfers associated with the storage system. In such systems, host controllers may provide the interface between the host computers and the storage system, and device controllers may be used to manage the transfer of data to and from an associated array of storage devices (e.g. disk drives). Often, the arrays may be accessed by multiple hosts and controllers. In addition, advanced storage systems, such as the SYMMETRIX® storage systems manufactured by EMC Corporation, generally include a global memory which typically shared by the controllers in the system. The memory may be used as a staging area (or cache) for the data transfers between the storage devices and the host computers and may provide a communications path which buffers data transfer between the various controllers. Various communication channels, such as busses, backplanes or networks, link the controllers to one another and the global memory, the host controllers to the host computers, and the disk controllers to the storage devices. Such systems are described, for example, in Yanai et al, U.S. Pat. No. 5,206,939 issued Apr. 27, 1993, (hereinafter “the '939 patent”), Yanai et al, U.S. Pat. No. 5,381,539 issued Jan. 10, 1995, (hereinafter “the '539 patent”), Vishlitzky et al, U.S. Pat. No. 5,592,492 issued Jan. 7, 1997, (hereinafter “the '492 patent”), Yanai et al, U.S. Pat. No. 5,664,144 issued Sept. 2, 1997 (hereinafter “the '144 patent”), and Vishlitzky et al, U.S. Pat. No. 5,787,473 issued Jul. 28, 1998, (hereinafter “the '473 patent”), all of which are herein incorporated in their entirety by reference. The systems described therein allow the controllers to act independently to perform different processing tasks and provide for distributed management of the global memory resources by the controllers. This high degree of parallelism permits improved efficiency in processing I/O tasks. Since each of the controllers may act independently, there may be contention for certain of the shared memory resources within the system. In these systems, the consistency of the data contained in some portions of global memory may be maintained by requiring each controller to lock those data structures which require consistency while it is performing any operations on them which are supposed to be atomic.
Since locking inherently reduces the parallelism of the system and puts a high load on system resources, locking procedures must be designed with care to preserve system efficiency. Adding features to the lock, such as queuing, lock override procedures, or multimodality can help to avoid some pitfalls of common lock protocols, such as processor starvation, deadlocks, livelocks and convoys. However, it is also known that, while many of these lock features have individual advantages, multifeatured lock management procedures are difficult to design and implement without unduly burdening system resources or inadvertently introducing pitfalls such as additional deadlock or starvation situations. For example, multimodal locks, which permit the requestor to identify the kind of resource access desired by the requestor and the degree of resource sharing which its transaction can tolerate, can be useful in improving system performance and avoiding deadlocks, but providing a lock override which is suitable for a multimodal lock is quite difficult. If, for example, one lock mode is set to allow unusually long transactions, a timeout set to accommodate normal transactions will cut the long ones off in midstream while a timeout set to accommodate the long transactions will allow failures occurring during normal transactions to go undetected for excessively long periods. Moreover, timeouts are competitive procedures which, in certain circumstances, undesirably offset the cooperative advantages of a queued lock. Because of the complexities introduced by multifeatured locks, it is desirable to validate features and modes which create particularly significant drains on system resources, such as long timeout modes, but introducing additional validation features can itself load system resources to the point where the system efficiency suffers.
Providing suitable procedures becomes especially difficult in complex multiprocessor systems which may contain a number of queued locks associated with different shared resources and where a requestor may have to progress through a number of lock request queues in turn in order to complete a process. In these systems, it is desirable that whatever procedure is implemented be fair, ensure that each requestor eventually obtains access to the lock whether or not all other requestors in the system are operating properly, and minimize the average waiting time for each requestor in the queue to improve system efficiency. Queued locks which implement a first-in-first-out (FIFO) protocol meet the fairness criteria because denied requests are queued in the order they are received. One such lock services procedure, often known as the “bakery” or “deli” algorithm, is described, for example, in “
Resource Allocation with Immunity to Limited Process Failure
”, Michael J. Fischer, Nancy A. Lynch, James E. Burns, and Alan Borodin, 20
th
Annual Symposium on Foundations of Computer Science
, San Juan, Puerto Rico, October '79, p 234-254; and “
Distributed FIFO Allocation of Identical Resources Using Small Shared Space”, ACM Transactions on Programming Languages and Systems
, January '89, 11(1): 90-114. When all requestors in the system are operating properly, the basic “deli” algorithm also meets the other criteria, but a protocol violation such as the failure of any processor in the lock request queue can lead to total system deadlock. However, in all complex multiprocessor systems, occasional protocol violations are inevitable, and the “deli” algorithm makes no provision either for detecting these through validation procedures or otherwise, or for handling them when they occur. Moreover, the basic “deli” lock is a unimodal lock.
A lock is needed which supports multiple locking modes and makes provision both for validation features to detect protocol violations and lock override procedures to manage the violations without unduly reducing system efficiency, and which also meets desirable design criteria for fairness, wait time minimization and guaranteed access.
SUMMARY OF THE INVENTION
In accordance with the present invention, a lock mechanism for managing shared resources in a data processing system is provided.
In accordance with the present invention, a method for providing queued locking and unlocking services for a shared resource is provided. The services include a cooperative lock override procedure. In one aspect, the locking services are multimodal and the cooperative lock override procedure is selectively associated with a lock mode.
In another aspec
EMC Corporation
Gunther John M.
Portka Gary
Wilson Penelope S.
LandOfFree
Cooperative lock override procedure does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Cooperative lock override procedure, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Cooperative lock override procedure will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3326192