J. Gray, R. A. Lorie, G. R. Putzolu, and I. L. Traiger. Granularity of locks and degrees of consistency in a shared data base. In IFIP Working Conference on Modelling in Data Base Management Systems, pages 365--394, 1976.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....the remote program can be kept weakly synchronized with the database. Moreover, if the machines time is synchronized [Mil95] then the remote program also knows the freshness of their data. Timestamp consistency control is very simple to implement in comparison to other consistency protocols [GLP75] and if the database is accessible, then the data is always available regardless of other failures in the system, whereas other protocols may deny access to ensure stricter consistency. Finally, we use timestamps to eliminate stale data. Stale data can occur because of failures or removals. The ....

Jim Gray, Raymond A. Lorie, and Gianfranco R. Putzulo. Granularity of locks and degrees of consistency in a shared database. In 1st International Conference on VLDB, pages 428--431, September 1975. Reprinted in Readings in Database Systems, 3rd edition.

....transaction management toolkit consists of several pieces: transactional IPC, a transaction manager, and a log manager. We will briefly describe each component of the toolkit in the remainder of this section. An earlier paper [10] t Following the terminology introduced by Gray et al. [6], the QuickSilver file system provides degree 2 consistency for file updates and degree 1 consistency on directories. presents the overall architecture of transaction management in QuickSilver and describes the toolkit in more detail. Transactional IPC. IPC in QuickSilver follows the ....

....two transactions from renaming a file to the same name, for example, but read locks are not required to read a directory. Thus, it is possible for clients to read directory entries that have been changed by a transaction that may later abort. According to the terminology introduced by Gray et al. [6], DFS provides degree 1 consistency for directories. The choice of locking policy for individual files is best illustrated by the example of a backup utility. If all files read by the backup program remained locked until the backup transaction ended, other programs that wanted to update those ....

J. N. Gray, R. A. Lorie, G. R. Putzolu, and I. L. Traiger. Granularity of locks and degrees of consistency in a shared data base. In G. M. Nijssen, editor, Modelling in Data Base Management Systems, pages 365-394. North Holland Publishing Company, 1976.

....any set of interleavings thus defined has a simple characterization, in terms of the absence of cycles in a particular dependency relation among transaction steps. This characterization ought to be useful in the design of concurrency control algorithms for multilevel atomicity. Other researchers [2, 5, 6, 7] have also noted that the usual notion of serializability needs to be weakened. In particular, reference [5] contains interesting preliminary work on specification and concurrency control design, for certain nonserializable interleavings. In fact, the multilevel atomicity of this paper is a ....

GRAY, J. N., LORIE, R. A., PUTZOLU, G. R., AND TRAIGER, L. I. Granularity of locks and degrees of consistency in a shared database. In Proc. IFIP Working Conf. on Modelling of Data Base Management Systems (Freudenstadt, Germany, Jan. 1976), pp. 695-723. Also in Modelling in Data Base Management Systems, G. M. Nijssen, Ed., Elsevier North-Holland, 19'16, pp. 365-395.

....objects synchronize the actions that access them by the use of locks. Every operation on an object is classified as a read or write, and an appropriate lock (read lock or write lock) must be obtained (automatically, by our system) before the operation can take place . Strict two phase locking ([Gray, et al. 1976]) is employed; locks are held until their action commits or aborts. The rules for acquiring locks are derived from conflicts between operations: A write operation conflicts with any other write or read operation. Without nesting, the rules are simple: many concurrent readers are allowed, but when ....

J. N. Gray, R. A. Lori, G. F. Putzolu, and I. L. Traiger. Granularity of Locks and Degrees of Consistency in a Shared Data Base. In G.M. Nijssen, editor, Modeling in Data Base Management Systems, pages 365--394, North Holland, Amsterdam, 1976. Also available as IBM Research Report RJ 1654/1706.

....but if this is not the case, then it should delegate the Index Manager of PHOS the management of the locking policy. 3. MGL protocol and the Expandable MGL technique 3. 1 Multigranularity Locking in ODMSs The first proposal of use of the MGL protocol appeared in the context of relational DMSs [Gray 76] In this context, the use of MGL aims at minimising the number of locks that a transaction needs to acquire when it accesses the set of tuples of a relation. To this purpose, a transaction can lock a granule in five different modes (S, X, IS, IX, SIX) In this paper we assume the reader s ....

Gray J., R. Lorie and G. Putzolu, "Granularity of locks and degrees of consistency in a shared database", IBM Res. Rep. RJ1654, IBM Research Laboratory, San Jose, CA, also in Modeling in Database Management Systems, Nijssen (ed.), North Holland, 1976.

....because agents outside the system may act based upon system output; though the system may be able to undo internal actions, in general the effects of output cannot be undone. We will discuss this point more later. This discussion of consistency is based mainly on the following works: EGLT76, Gray75, Gray78] Because external consistency is not as familiar a concept as internal consistency, let us discuss it in more detail. Suppose that a user is interacting with the system via a terminal. A possible external consistency requirement is that once the system has typed done on the terminal in ....

....write mode lock for an object, that action is permitted to read the object as well as to write it. At most one action at a time may 1. The motivation for using locking instead of timestamps was a desire to offer a contrast and alternative to Reed s timestamp based scheme [Reed78] 2. See [Gray78, Gray75] for a method that associates locks also with hierarchically organized groups o objects. 3. For now we are dcdbing mutual exclusion locks. Read write locking will be introduced a liUle lair. hold a lock in write mode; any number of actions may hold a lock in read mode concurrently. However, if ....

[Article contains additional citation context not shown here]

J. N. Gray, et. al., "Granularity of Locks and Degrees of Consistency in a Shared Data Base", IBM Research Report RJ1654, Sept. 1975.

....withdraws money from the same account accessed by transaction X, then that failing withdrawal might cause inconsistency. An approach to avoid inconsistency and, at the same time, allow for the interleaving of operations of different transactions, is to guarantee that the schedule is serializable [5]. The characteristic of serializable schedules is that they cause results in the database that are equivalent to results from a serial schedule. A popular kind of serializability is conflict serializability. A schedule is conflict serializable if it is conflict equivalent to a serial schedule, i. ....

....be inconsistent. Schedule C is not serial, but the order in which conflicting operations are processed is the same as in schedule A. They are conflict equivalent; schedule C is (conflict ) serializable. It can be demonstrated that serializable schedules produce correct results in the database [5]. In summary, schedule C produces correct results as if it was serial, and has the advantage of intercalating operations from different transactions, thus increasing concurrency and allowing for a better performance. Once students recognize serializability as an interesting property of schedules, ....

J. Gray, R. Lorie, and G. Putzulo, "Granularity of locks and degrees of consistency in a shared database", in G. Nijssen (ed.), Modelling in data base management systems, North-Holland, 1976.

....has to be provided. In fact, every practical DBMS supports such a hierarchy of typically 2, 3, or 4 levels; e.g. System R has a generic 4 level hierarchy: database segment relation tuple [Astrahan76] An appropriate hierarchical locking scheme was proposed for flat transactions in [Gray76]. Two key ideas allowed for the design of a scheme that could be adapted to a transaction s needs for either locking a few items using a fine lockable unit or locking larger sets of items with larger lock granules: A node R in a hierarchy can be locked explicitly. As a result, its entire ....

....A further refinement is the Share and Intention eXclusive mode (SIX) 19 which grants an S lock for the entire subtree to a transaction. In addition, it indicates the transaction s intention to request X locks explicitly for finer object granules later on. The following table taken from [Gray76] shows the compatibilities among request lock modes which derive from these semantics: For a comprehensive discussion of the precise effects of the lock modes and their compatibilities we refer the reader to the seminal work of J. Gray [Gray78] Basic Locking Rules for Object Hierarchies We ....

[Article contains additional citation context not shown here]

Gray, J.N., Lorie, R.A., Putzolu, G.R., Traiger, I.: Granularity of Locks and Degrees of Consistency in a Shared Data Base. Proc. IFIP Working Conf. on Modelling in Data Base Management Systems, North-Holland, 1976, 365-394.

.... However, a Full Global Index is costly to maintain due to remote updates (which incur communication as well as locking costs) and is also costly to use for complex queries because: Remote share (S) lock(s) must be acquired to assure serializability (i.e. repeatable read or degree 3 isolation [GrLo76] Otherwise, the retrieved record(s) have to be re certified for all index search conditions. Index ANDing ORing [MoHa90] for multiple index access) is expensive unless the participating Global Indexes a re stored at the same node and the record ID lists obtained from the indexes are short. ....

.... Global Index or Local Index can be recovered separately in a conventional way (e.g. as in System R [GrMc81] or ARIES [MoHa92, Moha99] We assume that the readers of this paper are familiar with the System R lock modes, the compatibility relationships amongst them and the durations of locking [GrLo76, MoLe92, Mo90b] We also assume that each Global or Local Index is implemented using a B tree. Besides maintaining the consistency between a Global Index and its Local Indexes, this protocol also exploits the following techniques to assure correct operation on each individual (Global or ....

Gray, J. , Lorie, R., Putzolu, F., and Traiger, I., Granularity of Locks and Degrees of Consistency in a Shared Data Base, Proc. IFIP Working Conference on Modelling of Database Management Systems , Freudenstadt, January 1976.

....aggregation hierarchies (in fact DAGs) However, many accesses in a KB are directed to a particular hierarchy, and not to the KB graph as a whole. These observations build the main idea of LARS. In LARS, the KB graph is partitioned into those three main hierarchies, and hierarchical lock schemes [8] are applied on each one of them. As a result, a minimization of the locks is obtained. In addition, the granule of lock to be accessed by a transaction is more precisely defined, allowing it to lock just the objects it really needs to access. Following these logical partitions, LARS provides ....

....be coped with by LARS. First, if the locks requested and granted give respect to the same set of objects (either C type vs. C type, or S type vs. S type, or A type vs. A type) then the compatibility matrix to be followed is the same of the Multigranularity Locks protocol known from the literature [8] (Table 3) The second situation with respect to the compatibility of the typed locks is the one where both are of different types (either C type vs. S type or A type , or S type vs. C type or A type , or A type vs. C type or S type ) In this case, the compatibility of the lock modes is not ....

[Article contains additional citation context not shown here]

Gray, J.N., Lorie, R.A., Putzolu, G.R., Traiger, I.: Granularity of Locks and Degrees of Consistency in a Shared Data Base. Proc. of the IFIP Working Conf. on Modelling in DBMSs, North-Holland, 1976.

....protect read operations, each transaction requests an S lock; to protect update operations, an X lock is necessary. Because databases often have multiple levels of granularity (i.e. the entire database, a single file, a single record, etc. these basic locks were extended to granularity locks [52]. Another reason for such locks is object composition (as Orion [47] A new lock type, called intent, represents the intention to acquire a lock at a finer level of granularity. A transaction s lock requests need to be augmented to take such granularity into account. For an operation on a data ....

....we must take care not to force already donated pages. Thus, implementing the Donate operation itself required some effort. In [95] the authors discuss other problems, related to recovery, that AL might introduce. The second problem reveals a subtle interaction between granularity locking [52] and AL. When scanning a file, Exodus acquires a file lock instead of acquiring a separate lock for each page in the file. The transaction scanning a file, however, cannot donate this file lock until it completes the scan. To work around this problem, we programmed a client application to mimic a ....

J. Gray, R. Lorie, and G. Putzolu. Granularity of locks and degrees of consistency in a shared database. In International Conference on Very Large Data Bases, pages 428--451. Morgan Kaufmann, 1975.

....all the sites in the DDBS, the following CC algorithm termed Send on Demand was proposed. The database correctness notion was that of conflict serializability [11, 12] To avoid rollbacks to the maximum extent, the consistency criteria of the highest degree (degree 3 as defined by Gray et al. in [11, 13]) was used. In the new scheme, every incoming transaction T broadcasts its access set information along with its arrival (at the head of line position) timestamp, TS(T) to all the sites in the DDBS. Every site receiving the broadcast message compiles this information to construct a claim queue ....

J. Gray, R. Lorie, G. Putzolu, and I. Traiger, "Granularity of Locks and Degrees of Consistency in a Shared Data Base," in Modelling in Data Base Management Systems (G. Nijssen, ed.), pp. 365--394, Amsterdam: North-Holland, 1976.

....items to synchronize transactions. It is sometimes desirable, however, to be able to access a set of data items as a single unit. Gray et al. presented a multiple granularity concurrency control protocol that aims to minimize the number of locks used while accessing sets of objects in a database [Gray et al. 75] In their model, Gray et al. organize data items in a tree where small items are nested within larger ones. Each non leaf item represents the data associated with its descendants. This is different from the tree protocol presented above in that the nodes of the tree do not represent the order of ....

Gray, J., Lorie R., and Putzolu, G. Granularity of Locks and Degrees of Consistency in a Shared Database. IBM Research Report RJ1654, IBM Research Laboratory, San Jose, CA, 1975.

....manager (LM) The TM knows which lock to request based on its knowledge about rules (e.g. which step in the rule execution is involved) and built in commands in Marvel. The LM decides whether the lock request should be accepted or rejected based on a multiple granularity locking (MGL) protocol [GLPT76] The details of the LM and the adaptation of the MGL protocol are presented in [BSKH92] and [Bar92] If the LM grants the lock request, the TM informs the transaction that it can go ahead and execute the operation. Once the TM has obtained a lock on an object, the lock cannot be released until ....

J. Gray, R. Lorie, G. Putzolu, and I. Traiger. Granularity of Locks and Degrees of Consistency in a Shared Database. In Modeling in Data Base Management Systems, pages 365--395, Amsterdam Holland, 1976. North Holland.

No context found.

J. Gray, R. Lorie, G. Putzolu and, I. Traiger, "Granularity of Locks and Degrees of Consistency in a Shared Data Base," in Readings in Database Systems, Second Edition, Chapter 3, Michael Stonebraker, Ed., Morgan Kaufmann 1994 (originally published in 1977).

....group of workstations) connected by a local area network. The replication mechanism presented is based on the state machine approach [4] and di ers from traditional replication mechanisms in that it does not handle replication using distributed transactional mechanisms, such as atomic commitment [5, 6]. The state machine approach was proposed as a general mechanism for dealing with replication, however no previous study has addressed its use in the domain of a cluster of database servers. Our Database State Machine is based on the deferred update technique. According to this technique, ....

J. N. Gray, R. Lorie, G. Putzolu, and I. Traiger, Readings in Database Systems, ch. 3, Granularity of Locks and Degrees of Consistency in a Shared Database. Morgan Kaufmann, 1994.

No context found.

J. Gray, R. A. Lorie, G. R. Putzolu, and I. L. Traiger. Granularity of locks and degrees of consistency in a shared data base. In IFIP Working Conference on Modelling in Data Base Management Systems, pages 365--394, 1976.

No context found.

J. N. Gray et al. Granularity of locks and degrees of consistency in a shared data base. In G. M. Nijssen, editor, Proc. of IFIP TC-2 Working Conference on Modelling in Data Base Management Systems, pages 1--29. North-Holland, 1976.

No context found.

J. Gray, et al. "Granularity of Locks and Degrees of Consistency in a Shared Database", Modeling in Database Management Systems, G.M. Nijssen ed., Jan 1976.

No context found.

J. Gray, R. Lorie, G. Putzolu, and I. Traiger, "Granularity of Locks and Degrees of Consistency in a Shared Database," Modeling in Data Base Management Systems, Amsterdam: Elsevier North-Holland, 1976.

No context found.

Gray, J. N., Lorie, R. A., Putzolu, G. R., and Traiger, I. L. Granularity of Locks and Degrees of Consistency in a Shared Data Base. IFIP Working Conference on Modeling on Data Base Management Systems. AFIPS Press, 1977.

No context found.

J. N. Gray, R. A. Lorie, G. R. Putzolu, and I. L. Traiger. Granularity of locks and degrees of consistency in a shared data base. In G. M. Nijssen, editor, Proseedings of the IFIP Working Conference on Modelling in Data Base Management Systems, pages 365--394, Amsterdam, 1976.

No context found.

J. Gray, R.A. Lorie, and G.R. Putzolu et al. Granularity of Locks and Degrees of Consistency in a Shared Data Base. IFIP Working Conference on Modeling in Data Base Management Systems 1976: 365394.

No context found.

J.N. Gray, R.A. Lorie, G.R. Putzolu, and I.L. Traiger. Granularity of locks and degrees of consistency in a shared data base. In Modelling in Data Base Management Systems. North-Holland, 1976.

No context found.

J. Gray, R. Lorie, G. Putzolu, I. Traiger. Granularity of Locks and Degrees of Consistency in a Shared Data Base. IFIP Working Conf on Modeling of Data Base Management Systems. 1976

First 50 documents  Next 50

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC