R. K. Abbott and H. Garcia-Molina. Scheduling realtime transactions with disk resident data. In Proceedings of Very Large Database Conference, pages 385-- 396, 1989.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....efficient garbage collection distinguish our work. Our work on making read only transactions completely non blocking and their execution times more predictable is complementary to the time cognizant transaction processing schemes proposed in the real time database systems literature [SZ88, AGM88, AGM89, HCL90b, HCL90a, HSRT89, HSRT91, HCL93, Ram93] Our use of main memory technology and version based concurrency control schemes enable transaction running times to be estimated more accurately. Consequently, scheduling algorithms can generate better schedules, that is, schedules in which more ....

R. Abbott and H. Garcia-Molina. Scheduling real-time transactions with disk-resident data. In Procs. of the International Conf. on Very Large Databases, 1989.

....contract number E30602 97 2 0287 and in part under contract number F30602 96 1 0160. Mr. Molano was funded by a research grant from the Community of Madrid and by the National R D Program of Spain under contracts TIC96 0982 and TIC97 0438. one type (namely network packets) Work described in [1, 2, 7, 8] study various aspects of real time disks and filesystems. The approach used in [1] is a variation of the Scan algorithm in which the SCAN direction changes towards the Earliest Deadline First request only if the deadline is considered to be met. At any scheduling point the request with the ....

....Molano was funded by a research grant from the Community of Madrid and by the National R D Program of Spain under contracts TIC96 0982 and TIC97 0438. one type (namely network packets) Work described in [1, 2, 7, 8] study various aspects of real time disks and filesystems. The approach used in [1] is a variation of the Scan algorithm in which the SCAN direction changes towards the Earliest Deadline First request only if the deadline is considered to be met. At any scheduling point the request with the Earliest Feasible deadline is chosen for service and the disk head scans towards it. Chen ....

[Article contains additional citation context not shown here]

R. Abbott and H. Garcia-Molina. Scheduling Real-Time Transactions with Disk Resident Data X Server. Tech. Rept. CS-TR-207-89, Department of Computer Science, Princeton University, Feb., 1989.

....In this algorithm, the request with the earliest feasible deadline is chosen as the target and determines the scanning direction. A deadline is feasible if it is estimated that it can be met. If there is no request with a feasible deadline, then simply the closest request is serviced. In [2], Abbott and Garcia Molina studied an IO architecture that handles read requests and write requests differently. The architecture assumes that write operations of a transaction are always performed after the transaction has committed. While a read request is assigned a priority based on the timing ....

....(Priority Hints) was shown to perform as well as Priority DBMIN and better than Priority LRU. Abbott and Garcia Molina [3] presented and evaluated two new buffer management techniques to be used in scheduling IO requests with deadlines. On the basis of the IO architecture they provided in [2] (as discussed in the preceding section) read and write requests are treated separately by the proposed techniques. Read requests are assumed to be issued by uncommitted transactions and receive service in accordance with the timing constraints of the transactions that issued them. Write ....

[Article contains additional citation context not shown here]

R. Abbott, H. Garcia-Molina `Scheduling Real-Time Transactions with Disk Resident Data', Proceedings of the 15th International Conference on Very Large Data Bases, 1989, pp.385-396.

....access requests of transactions. An extensive exploration of the issues in RTDBSs is provided in [32] The transaction scheduling problem in RTDBSs has been addressed by a number of recent studies. The first attempt to evaluate the performance of scheduling algorithms in RTDBSs was provided in [1, 2]. Abbott and Garcia Molina described and evaluated through simulation a group of real time scheduling policies based on enforcing data consistency by using a two phase locking concurrency control mechanism. An extended version of their work appeared recently in [4] In [3] they provided a study ....

....tr length (i.e. the number of data items accessed by each transaction) The mean interarrival time value (iat) was fixed at 400 msec and the same value was assumed for the maximum transaction population (i.e. 50) with each tr length value considered. The range of tr length values employed was [2,10], which corresponds to a database access ratio of 0.05 to 0.25. The results are displayed in Figure 6 for protocols PI and OPT. The reaction of protocols PB and CP to the change in the database access ratio was similar to that of protocol PI, thus, PI was selected as representative for the locking ....

R.Abbott, H.Garcia-Molina, "Scheduling Real-Time Transactions with Disk Resident Data", 15th International Conference on Very Large Data Bases, 1989, pp.385-396.

....usually are considered to be trade secrets. This section will discuss a few success stories. 4. 1 Disks Disk Scheduling Numerous studies have investigated the performance of disk I O scheduling algorithms involving command queuing (e.g. 18, 51, 52, 63, 64, 65] real time applications (e.g. [66, 67, 68, 69]) and database applications (e.g. 70] Event driven simulation studies of disk scheduling algorithms that use rotational position information appear in [52] these studies were used by Hewlett Packard s Disk Memory Division in the design of products. Disk Cache Performance evaluation methods ....

R. Abbott and H. Garcia-Molina, \Scheduling real-time transactions with disk resident data," in Proceedings of 15th International Conference on Very Large Data Bases, (Amsterdam, The Netherlands), pp. 385-396, Aug. 1989.

....immediately the transaction should wait until it will become possible. Most of pessimistic algorithms are based on locks. The classical pessimistic algorithm is the widely used two phase locking (2PL) 5] Real time systems usually use modified versions of 2PL, such as 2PL HP, 2PL WP, 2PL PC [1, 2, 13]. These protocol modifications take into consideration real time aspects, such as transaction priorities. According to 2PL HP, conflict situations are resolved in favor of the transaction with higher priority. When a transaction requests a lock on an object held by an other transaction in a ....

R. Abbott and H. Garcia-Molina. Scheduling real-time transactions with disk resident data. In Procedings of the 15th VLDB Conference, August 1989.

....in a real time database system are usually associated with deadlines. Any deadline violation of a hard real time transaction may result in a catastrophe, while deadline violations of soft real time transactions might only cause a degraded level of system performance without a total system failure [AG89, HCL92, ORP96, YWLS92]. In the past two decades, researchers have proposed various efficient real time concurrency control techniques to either reduce the number of deadline violations for soft (and firm 1 ) real time transactions or guarantee the deadlines of hard real time transactions [AG92, BN96, HCL90, HS99, ....

....These factors are detrimental to satisfying the transaction deadlines [CSKT91] On the other hand, the probability of cache hit can significantly affect the impact of the disk scheduling on the performance of the protocols. If a deadline cognizant technique is adopted to schedule disk requests [AG89], requests from transactions with earlier deadlines may have a higher chance of being served first. The disk thus acts (indirectly) as an agent that controls the progress of the transactions, and the waiting time of non real time transactions will be much longer than that of the soft real time ....

Abbott, R. and H. Garcia-Molina, "Scheduling Real-time Transactions with Disk Resident Data", in Proceedings of the 15 th VLDB Conference, pp. 356-396, 1989.

.... Pu92, Rajk89, Sha88, Son90, Son92, Song90, Stan88, Vrbs88] The subjects addressed in these papers include the modeling of real time transactions and real time database systems [Abbo88, Buch89, Care89, Daya88, Hari90, Huan89, Kort90, Liu88, Sha88, Stan88] scheduling of real time transactions [Abbo88, Abbo89, Huan89, Stan88, Sha88], concurrency control and data conflict resolution [Abbo88, Abbo89, Agra92, Buch89, Hari90, Hari90b, Huan89, Huan91b, Huan91c, Lee93, Lee93b, Lee93c, Lee93d, Lin90, Sha88, Son90, Son90b, Song90, You93] processing of queries with real time constraints [Hou89] buffer management [Care89, Huan90] ....

.... these papers include the modeling of real time transactions and real time database systems [Abbo88, Buch89, Care89, Daya88, Hari90, Huan89, Kort90, Liu88, Sha88, Stan88] scheduling of real time transactions [Abbo88, Abbo89, Huan89, Stan88, Sha88] concurrency control and data conflict resolution [Abbo88, Abbo89, Agra92, Buch89, Hari90, Hari90b, Huan89, Huan91b, Huan91c, Lee93, Lee93b, Lee93c, Lee93d, Lin90, Sha88, Son90, Son90b, Song90, You93], processing of queries with real time constraints [Hou89] buffer management [Care89, Huan90] and I O scheduling [Abbo90, Care89, Chen91] A number of papers on related issues have also appeared. These include work on a protocol for timed atomic commitment [Davi91] fast recovery protocols for ....

[Article contains additional citation context not shown here]

Abbott, R. and H. Garcia-Molina, "Scheduling Real-Time Transactions with Disk Resident Data," Proceedings of the 15th VLDB Conference, August 1989.

....of ECA rules. In particular a lot of work has been done on semantics of event and rule specification and evaluation, as well as coupling modes between conditions and actions. None of this work considers a real time context. The pioneering work in RTDBS performance evaluation was reported in [1, 2, 3], where the authors simulated a number of CC protocols based on the two phase locking algorithm. However, this work was not examined in an active context. A subset of notable RTDBS work [23, 22, 25, 29, 44, 5, 14] Again, all this work has been performed without considering the effects of ....

R. Abbott and H. Garcia-Molina. Scheduling real-time transactions with disk resident data. In Proceedings of the 15th VLDB, 1989.

....protocols designed for a single type of transactions. For example, the well known priority ceiling protocol (PCP) Sha et al. 1990; Sha et al. 1991) designed for HRT) requires a static system that is usually not true for a system with SRT. The Higher Priority Two Phase Locking Protocol (HP 2PL) (Abbott and Garcia Molina, 1989; Abbott and Garcia Molina, 1992) which is designed for SRT, cannot guarantee the deadlines of HRT or even be suitable to a RTDBS that consists of mixed SRT and non real time transactions (NRT) The response times of NRT may be greatly affected because of the restart policy of HP 2PL. A NRT may ....

....results are provided in Section 5. Section 6 is the conclusion. 2 Real time Concurrency Control Protocols 2. 1 Real time Concurrency Control Protocols for Single Type of Transactions Real time concurrency control protocols are often extended from traditional concurrency control protocols (Abbott and Garcia Molina, 1989; Abbott and Garcia Molina, 1992; Haritsa et al. 1990; Haritsa et al. 1992; Huang et al. 1992; Sha et al. 1990; Sha et al. 1991; Ulusoy and Buchmann, 1998) Most lock based real time concurrency control protocols follow from the two phase locking scheme (2PL) Bernstein et al. 1987) For ....

[Article contains additional citation context not shown here]

Abbott, R. and H. Garcia-Molina. 1989. Scheduling Real-time Transactions with Disk Resident Data. In: Proceedings of the 15th VLDB Conference, pp. 356-396.

....may cause aborts or delays of indeterminate length. There has been extensive study of real time systems [Sta88] Formal aspects of such systems have been examined from the standpoints of scheduling (e.g. HMR 89] and verification [JM86] In the context of real time databases, AGM88, AGM89] consider alternative queuing disciplines with lock based concurrency control of real time transactions, and use simulation results to compare 2 these techniques. The research in [SRL88] proposes concurrency control techniques for distributed real time systems based on a partitioning of data, ....

R. Abbott and H. Garcia-Molina. Scheduling real time transactions with disk resident data. In Proceedings of the Fifteenth International Conference on Very Large Databases, Amsterdam, pages 385--396, 1989.

.... systems; Database management; Architecture; Intelligent memory 1 Introduction Applications that require real time functionality are difficult to implement using standard hardware and software systems [Sta88] This is also the case with a real time database (RTDB) management system [AGM88, AGM89, BMHD89, C 89, DLW90, HCL90, KSS90, Sin88, Son88, SZ88] which straddles the areas of a high performance database management system (DBMS) and a real time system (RTS) A general RTDB application may be regarded as a reactive process control system [Sin88] The world external to the computer ....

R. Abbott and H. Garcia-Molina. Scheduling real time transactions with disk resident data. In Proceedings of the Fifteenth International Conference on Very Large Databases, Amsterdam, pages 385--396, 1989.

....from the norm in that we do not directly address the question of time constrained scheduling at each local site. Instead, we consider issues pertinent to the distributed environment where the local scheduling is handled through the use of methods that have been previously explored [AGM88, AGM89, BMHD89, C 89, HCL90b, HCL90a, KSS90] Thus, a transaction that accesses data at several sites, does so by the use of a subtransaction at each concerned site, and each such subtransaction is subject to the local concurrency control mechanism at its particular site. We adopt such an approach ....

....trends and makes decisions regarding the sale or buying of options, and sends these to the sites to effect the transactions. The need for a coordinated set of actions at the different sites is evident. Also, the need for fast, real time responses is necessary to exploit the trends in the market [AGM89] We assume that there is a need to transfer money from one site to another, or to audit the total amount of money that exists in accounts managed at more than one site. Such systems are usually designed using a transaction processing paradigm. Consider a global transaction T 1 that transfers ....

R. Abbott and H. Garcia-Molina. Scheduling real time transactions with disk resident data. In Proceedings of the Fifteenth International Conference on Very Large Databases, Amsterdam, pages 385--396, 1989.

....transaction wakeup, deadlock, buffer management, and disk I O scheduling [21, 23, 25, 26, 50, 100] Each of these algorithms or protocols should directly address the real time constraints. To date, work on real time databases has investigated a centralized, secondary storage real time database [1, 2, 3]. As is usually required in traditional database systems, work so far has required that all the real time transaction operations maintain data consistency as defined by serializability. Serializability may be relaxed in some real time database systems, depending on the application environment and ....

R. Abbott and H. Garcia-Molina, "Scheduling Real-Time Transactions with Disk Resident Data," Proceedings of the 15th VLDB Conference, 1989.

....0.1, is used to determine whether a disk access is required. CPU scheduling is an important issue when dealing with real time transaction systems. We chose the following algorithm. ffl CPU Scheduling The earliest deadline (ED) algorithm is used for CPU scheduling. As mentioned in the literature [2], a major drawback of this algorithm is that it may assign the highest priority to a transaction that has already missed or is about to miss its deadline. To partially overcome this weakness, we propose a modified ED algorithm which, instead of scheduling according to a transaction s deadline, ....

Abbott, R. and Garcia-Molina, H., "Scheduling Real-Time Transactions with Disk Resident Data," A Technical Report, CS-TR-207-89, Princeton University, Feb. 1989.

....is known to be NP complete (see [17] and [36] Comprehensive surveys on real time databases can be found in [2] and [33] Our scheduling approach will follow some of the ideas of [2] especially using soft real time. Special scheduling algorithms for real time databases are e.g. detailed in [1], where earliest deadline, least slack and high priority strategies are studied. Another approach considers static and dynamic costs associated with tasks and transactions, see [23] Our approach uses domain specific knowledge about tasks and agents, particularly human users, for heuristic ....

R. Abbott and H. Garcia-Molina. Scheduling real--time transactions with disk--resident data. In Proc. Intl. Conf. on Very Large Data Bases, pages 385--396, Amsterdam, Netherland, 1989.

....time critical scheduling methods to observe the timing requirements of transactions [30] A considerable amount of RTDBS research has been devoted to performance evaluation of time cognizant concurrency control protocols. However, the performance studies were either based on simulation (e.g. [1, 2, 3, 10, 11, 20, 23, 27, 28, 29]) or carried out on a RTDBS testbed (e.g. 14, 15] To the best of our knowledge, no analytic performance study has been reported so far involving the evaluation of concurrency control protocols in RTDBSs, which is the main contribution of this paper. 1 The behavior of concurrency control ....

R. Abbott, H. Garcia-Molina `Scheduling Real-Time Transactions with Disk Resident Data', 15th International Conference on Very Large Data Bases, pp.385-396, 1989.

....to disk and redoing work of committed transactions that has been lost due to crashes. 5. Proper logging techniques and the proper placement of logs in the appropriate level of storage. The transaction scheduling and concurrency control aspects of real time databases have been studied in detail in [1, 9, 10, 11]. Most investigations have focused on the processing of transactions with soft deadlines by adopting priority assignment policies and conflict resolution mechanisms that explicitly take time into account. The most common results have been the development of various time cognizant extensions of two ....

R. Abbott and H. Garcia-Molina, "Scheduling Real-Time Transactions with Disk Resident Data", Proceedings of the 15th International Conference on Very Large Databases, 1989.

....database systems, scheduling and concurrency control used for conventional systems should be re evaluated and might need to be extended for RTDBS. Several research efforts have been carried out in evaluating the performance of concurrency control and the impact of real time constraints in RTDBS [AGM88, AGM89, HSTR89, Son91]. Buchmann et al. presented a framework for integrating real time scheduling and concurrency control and a summary of real time concurrency control algorithms [BMHD89] Several priority ceiling based protocols were proposed; these protocols provide bounded blocking times imposed by accessing data ....

R. Abbott and H. Garcia-Molina. Scheduling real-time transactions with disk resident data. In Proceedings of the 15th on VLDB Conference, pages 385--396, 1989.

....control methods affects the scheduling policy. Some work carried out studying the effects are [1] 2] 45] and [42] For distributed systems concurrency control has been studied in [14] The handling of resource requirements with respect to database transaction scheduling have been addressed in [3], 13] 18] 39] 40] 41] and [44] There are three good surveys covering additional aspects of research not considered here ( 38] 23] 29] 3 Active Real Time Database Systems 3.1 Applications Both active and real time capabilities have been considered as important in order to support ....

R. Abbott and H. Garcia-Molina. Scheduling real-time transactions with disk resident data. In Proceedings 15th International Conference on Very Large Data Bases, Los Angeles, 1989.

....studies related to transaction scheduling in RTDBSs; however, these studies were not specifically concerned with the performance of underlying transaction processing architectures. The first attempt to evaluate the performance of transaction scheduling algorithms in RTDBSs was provided in [Abbott and Garcia Molina 1988, 1989]. The authors described and evaluated through simulation a group of real time scheduling policies based on enforcing data consistency by using a two phase locking concurrency control mechanism. Huang et al. developed a new lock based concurrency control protocol by combining some existing schemes ....

R. Abbott, H. Garcia-Molina, Scheduling Real-Time Transactions with Disk Resident Data, 15th International Conference on Very Large Data Bases, 385-396.

....of dynamically arriving transactions that have deadlines. It assumes that transactions operate on a disk resident database, where only a small fraction of accessed data can be cached in memory. Although considerable research has been done recently on processing transactions with time constraints [1, 2, 4, 10, 13, 14, 15, 16, 20] for the most part, this work has not directly tackled the problems introduced by the various sources of unpredictability (discussed in Section 2) Whereas concurrency control protocols have been proposed to execute transactions taking their time constraints into account, most of them do not ....

R. Abbott and H. Garcia-Molina, "Scheduling Real-Time Transactions with Disk Resident Data," Proceedings of the 15th VLDB Conference, 1989.

.... generally considered that hard deadline RTDBSs are infeasible since it is difficult to determine a priori the computation time and execution pattern of a transaction [Abbo88, Stan88] The studies can be divided into two general groups those that treat all transactions as being equally important [Abbo88, Abbo89, Abbo90, Hari90a, Hari90b, Huan90a], and those that incorporate the notion of transactions having different values [Huan89, Huan90b] A brief summary of the studies in these two groups is presented in the remainder of this section. The problem of scheduling transactions in a RTDBS was first addressed by Abbott and Garcia Molina ....

....and those that incorporate the notion of transactions having different values [Huan89, Huan90b] A brief summary of the studies in these two groups is presented in the remainder of this section. The problem of scheduling transactions in a RTDBS was first addressed by Abbott and Garcia Molina [Abbo88, Abbo89]. Their work focused on evaluating the performance of various real time scheduling policies, all of which policies enforced data consistency by using a two phase locking protocol as the underlying concurrency control mechanism. In [Hari90a] the focus was shifted to studying the performance of ....

[Article contains additional citation context not shown here]

Abbott, R., and Garcia-Molina, H., "Scheduling Real-Time Transactions with Disk Resident Data," Proc. of the 15th Int. Conference on Very Large Database Systems, August 1989.

....with Soft Deadlines Transaction Priority Assignment Policies: Priorities and values for real time transactions are used for conflict resolution and CPU scheduling. The literature contains various transaction priority or value assignment algorithms [26, 75, 76, 86, 91, 134] and their evaluations [3, 1, 86]. Some of these policies are Earliest deadline first, Highest value first, Least slack time first (where slack time is the maximum amount of time that a transaction can spend without executing, and still complete within its deadline) Fixed priority with a priority ceiling, and ....

....the performance of the transaction scheduling algorithms. Concurrency Control Techniques with Serializability: Concurrency control techniques for real time databases that use serializability as the correctness criteria include lock based protocols such as two phase locking and its variants [3, 1, 7, 60, 86, 88, 179, 180, 193, 214, 216], optimistic concurrency control protocols [74, 78, 88, 121] and timestamp ordering protocols [133, 195, 215] Using any of these techniques, conflicts between two real time transactions or between one real time transaction and a set of real time transactions are detected. For lock based ....

[Article contains additional citation context not shown here]

Abbott, R., and Garcia-Molina, H., "Scheduling Real-Time Transactions with Disk resident Data," Proceedings of the International Conference on Very Large Databases, Aug. 1989.

....superior to either SCAN or SSTF. 2.2 New Disk Scheduling Algorithms The additional requirements introduced by applications, such as digital multimedia and database systems, make traditional disk scheduling algorithms described above inappropriate. Many new scheduling algorithms have been developed[3, 17, 2, 14, 6, 21]. 2.2.1 Multimedia Specific Algorithms EDF The Earliest Deadline First algorithm is an analog of FCFS. Disk requests are ordered according to deadlines and the request with the earliest deadline is serviced first. EDF has similar expected seek time as FCFS, since no positional information is used ....

R.K. Abbott and H. Garcia-Molina. Scheduling real-time transactions with disk resident data. In Proc. of the 15th International Conference on Very Large Data Bases, Amsterdam, 1989.

.... transactions have different values, both deadline and value must be considered [25] For the purpose of conflict resolution in real time database systems, various time cognizant extensions of two phase locking, optimistic, and timestamp based protocols have been proposed in the literature [1, 2, 9, 20, 25, 27, 26, 34, 47, 49]. These are discussed below. In the context of two phase locking, when a transaction requests a lock that is currently held by another transaction we must take into account the characteristics of the transactions involved in the conflict. Considerations involved in conflict resolution are the ....

R. Abbott and H. Garcia-Molina, "Scheduling Real-Time Transactions with Disk Resident Data," Proceedings of the 15th VLDB Conference, 1989.

....of ECA rules. In particular a lot of work has been done on semantics of event and rule specification and evaluation, as well as coupling modes between conditions and actions. None of this work considers a real time context. The pioneering work in RTDBS performance evaluation was reported in [1, 2, 3], where the authors simulated a number of CC protocols based on the two phase locking algorithm. However, this work was not examined in an active context. In [23] the problem of assigning deadlines to subtransactions 1 If temperature goes above 1000 degrees, then reduce pressure by 50 psi ....

....transactions is known in advance. In this work, we do not make any such assumption. Furthermore, much research has also been devoted to designing concurrency control (CC) mechanisms geared towards improving the timeliness of transaction processing and their subsequent performance evaluation [1, 2, 3, 20, 19, 22, 21, 29, 28, 41, 6, 27]. Again, all this work has been performed without considering the effects of triggering. An important result that we draw upon in this paper is reported in [20, 19] In this set of important studies, Haritsa et al. showed that in firm or hard real time scenarios (i.e. where late transactions are ....

[Article contains additional citation context not shown here]

R. Abbott and H. Garcia-Molina. Scheduling real-time transactions with disk resident data. In Proceedings of the 15th VLDB, 1989.

....on temporal query processing, and [15] which is one of the very few papers that we are aware of, on temporal transaction processing. RTDBSs have seen substantial research effort as well in recent years. Much of this effort has been focussed towards developing high performance scheduling algorithms [1, 2, 14, 21, 25, 31, 37] as well as concurrency control algorithms [6, 22, 24, 26, 27, 33] Typically, performance has been characterized as the ability to reduce transaction tardiness. None of this work has been performed with temporal consistency in mind. Even though not much is reported on the confluence of temporal ....

....(number of CPUs and disks) The basic parameter settings for our experiments are shown in Table 4 below. Parameter Value MTD 60 sec ProcCPU 10ms ProcDisk 20ms ContextSwitch 1ms DBSize 10000 pages SensorFrac 0:9 ItemsPerPage 1 NumSDItems 900 UpdateRate 3 WriteProb 0:25 SizeInterval [1,15] SRInterval [2,6] RQTSInterval [Current time Gamma 5 Theta MTD, Current time] ATIFactor 5 ThresholdMR 0.2 ThresholdFSR 0.3 Delta 1 SampleBatch 100 Table 4: Parameters Settings for Our Experiments 7.1 UPUM We first discuss the results of the experiments under the UPUM model, i.e. updates ....

R. Abbott and H. Garcia-Molina. Scheduling real-time transactions with disk resident data. In Proceedings of the 15th VLDB, 1989.

....of ECA rules. In particular a lot of work has been done on semantics of event and rule specification and evaluation, as well as coupling modes between conditions and actions. None of this work considers a real time context. The pioneering work in RTDBS performance evaluation was reported in [1, 2, 3], where the authors simulated a number of CC protocols based on the two phase locking algorithm. However, this work was not examined in an active context. A subset of notable RTDBS work [24, 23, 26, 29, 42, 4, 13] Again, all this work has been performed without considering the effects of ....

R. Abbott and H. Garcia-Molina. Scheduling real-time transactions with disk resident data. In Proceedings of the 15th VLDB, 1989.

....The check that needs to be undertaken is T currenttime T executiontime T deadline ; if the condition holds, the deadline is feasible. If there are many tasks with feasible deadlines, then the task with the earliest deadline will be assigned the highest priority (Abbott and Garcia Molina, 1988a; Abbott and Garcia Molina, 1989). The problem with the method is that a task can only get high priority when it is close to its deadline (Hong et al. 1992) and it can also start to behave unpredictably under heavy workload and when tasks have begun to miss their deadlines. The advantage is that the algorithm is known to be ....

Abbott, R. and Garcia-Molina, H. (1989). Scheduling real-time transactions with disk resident data. In Proceedings 15th International Conference on Very Large Data Bases, Los Angeles.

....information in the resolution of data conflicts, that is, resolve data conflicts always in favor of the higher priority transaction. The problem of scheduling transactions in an RTDBS with the objective of minimizing the percentage of transactions missing its deadline was first addressed in [Abbo88, Abbo89]. The work proposed new policies for priority assignment and eligibility test in RTDBSs, and evaluated the performance of various scheduling policies through simulation. A group of concurrency control protocols for RTDBSs using two phase locking as the underlying mechanism was also proposed and ....

....new policies for priority assignment and eligibility test in RTDBSs, and evaluated the performance of various scheduling policies through simulation. A group of concurrency control protocols for RTDBSs using two phase locking as the underlying mechanism was also proposed and evaluated. Studies in [Abbo89, Huan91b, Sha88, Sha91] addressed the problem of priority inversion in the context of RTDBSs where locking and priority driven preemptive scheduling for CPU scheduling are integrated together. A priority inversion occurs when a higher priority transaction must wait for the execution of lower priority transaction(s) ....

[Article contains additional citation context not shown here]

Abbott, R. and H. Garcia-Molina, "Scheduling Real-Time Transactions with Disk Resident Data," Proceedings of the 15th VLDB Conference, Aug. 1989.

....values to transactions and where the goal is to maximize the total value of the in time transactions. 1. INTRODUCTION A Real Time Database System (RTDBS) is a transaction processing system that is designed to handle transactions with completion deadlines. Several previous RTDBS studies (e.g. [Abbo88, Abbo89]) have addressed the issue of scheduling transactions with the objective of minimizing the number of late transactions. A common observation of these studies has been that assigning priorities to transactions according to an Earliest Deadline [Liu73] policy minimizes the number of late ....

Abbott, R., and Garcia-Molina, H., "Scheduling Real-Time Transactions with Disk Resident Data," Proc. of the 15th Int. Conf. on Very Large Database Systems, Aug. 1989.

....scheduling problem in RTDBS s has been addressed by a number of recent studies. Abbott and Garcia Molina described and evaluated through simulation a group of real time scheduling policies based on enforcing data consistency by using a two phase locking concurrency control mechanism [4] [6], 7] Two of the concurrency control protocols evaluated were priority based locking (PB) and priority inheritance (PI) In [8] they provided a study of various algorithms for scheduling disk requests with deadlines. The priority ceiling protocol (PC) was presented in [1] and [3] and the ....

....that the performance obtained by employing realtime policies based on the PB concurrency control protocol was better, in general, than that obtained in a nonreal time transaction processing system. Abbott and Garcia Molina also found that both protocols PI and PB perform better than protocol AB [6]. Their results indicated that no protocol is the best under all conditions; the comparative performance of the protocols PI and PB depends on some other factors they considered, such as the type of load, and the priority policy. Under continuous and steady load, the performance of protocol PI was ....

[Article contains additional citation context not shown here]

R. Abbott, H. Garcia-Molina. Scheduling Real-Time Transactions with Disk Resident Data. Proc. 15th Int. Conf. on Very Large Data Bases, 385-396 (1989).

....Most previous RTDBS studies have attempted to address the issue of scheduling transactions by assigning priorities. Moreover, in the context of priority based transaction scheduling in RTDBSs, most performance studies have adopted the Earliest Deadline (ED) principle for priority assignment [15, 1, 8, 11]. The ED policy has been shown to minimize number of missed transactions in lightly to moderately loaded systems (i.e. underloaded systems 1 ) 2] However, under overload conditions, transactions gain high priority under ED policy only when they are close to their deadlines and thus may not ....

....in section 5.3.2. A comparison of the computational overheads of AAP and AEVD is presented in section 5.4. Table 3 shows the parameter settings used for the baseline model. In Workload Parameter Value Resource Parameter Value DBSize 1000 pages NumCPUs 8 WriteProb 0 NumDisks 16 SizeInterval [1,30] ProcCPU 10ms SRInterval [2.0,6.0] ProcDisk 20ms Table 3: Parameter Settings for the Baseline Model 4 We also simulated the performance of ED to add perspective to our experiments. All our baseline model results include ED performance 0 10 20 30 40 50 60 70 80 25 30 35 40 45 50 55 60 65 70 ....

[Article contains additional citation context not shown here]

R. Abbott and H. Garcia-Molina. Scheduling real-time transactions with disk resident data. In Proceedings of the 15th VLDB, 1989.

....is not much work at all on the synthesis of these fields, which is what this paper proposes. There has been substantial research efforts devoted to Real Time database Systems (RTDBSs) in recent years. Much of this effort has been focussed towards developing high performance scheduling algorithms [1, 2, 22, 25, 31, 12, 34] as well as concurrency control algorithms [21, 26, 27, 24, 23, 32] For a nice discussion on the requirements of RTDBSs see [20] and for a nice survey of recent work see [48] Typically, performance has been characterized as the ability to reduce transaction tardiness. None of this work has been ....

....to the highest QT arrival rate at which a given algorithm can still process all QTs within DRT time units. Parameter Value MTD 60 sec ProcCPU 10ms ProcDisk 20ms NumCPU 2 NumDisk 4 NumSensor 250 DBSize 1000 pages SensorFrac 0:9 ItemsPerPage 1 NumSDItems 900 SRTLoadFac . 25 SizeInterval [1,15] SRInterval [2,6] RQTSInterval [Current time Gamma 5 Theta MTD, Current time] ATIFactor 5 ThresholdMR 5 ffi 1 SampleBatch 100 DRT 2.5 sec Table 2: Parameters Settings for Our Experiments 0 20 40 60 80 100 0 2 4 6 8 10 12 14 16 18 20 Arrival Rate, QTs per second [A] SDH SRTF NP QTF 0 2 4 ....

R. Abbott and H. Garcia-Molina. Scheduling real-time transactions with disk resident data. In Proceedings of the 15th VLDB, 1989.

....attempts to resolve data contentions by aborting transactions that conflict with a transaction having a smaller timestamp. A locking protocol, which is widely accepted and well performed mechanism under conventional databases, has been intensively studied for a real time database system. 2PL HP[1 3] is the real time locking protocol that favors higher priority transactions in resolving data conflicts. It, however, has the problem of requiring a prior knowledge of transactions. The research in [5] added real time constraints to ordered shared locking protocol that has been proposed as a ....

....of unnecessary restarts which degrade the performance of optimistic algorithms. The optimistic algorithm was designed to avoid unnecessary restarts by dynamically adjusting serialization order among concurrent active transactions. The study showed that optimistic algorithms including those in [1,3] outperform 2PL HP by discarding tardy transactions from the system. However, optimistic algorithms as in 2PL HP require information about running transactions such as the read write data set and the timestamp interval. There has been a concurrency control algorithm that combines the optimistic ....

R. Abbott and H. Garcia-Molina. Scheduling real-time transactions with disk resident data. In Prooceedings of the 15th International Conference on Very Large Data Bases, pages 385--396, 1989.

....algorithms attempt to resolve data contentions by aborting transactions that conflict with the validating transaction. A locking protocol, which is widely used and wellperformed mechanism under conventional database environment, has been intensively studied for a real time database system. 2PL HP[1,2] is a real time locking protocol that favors higher priority transactions in resolving data conflicts. It, however, has the problem of requiring a prior knowledge of transactions. The work in [4] showed that the optimistic strategy can outperform two phase locking(2PL) under firm real time ....

R. Abbott and H. Garcia-Molina. Scheduling real-time transactions with disk resident data. In Prooceedings of the 15th International Conference on Very Large Data Bases, pages 385--396, 1989.

....a priority based on its timing constraint. Transaction priorities are used in resolving data and resource conflicts among concurrently executing transactions. The papers published on priority assignment methods and their performance evaluation can be listed as: Abbott Garcia Molina 1988] [Abbott Garcia Molina 1989], Abbott Garcia Molina 1992] Biyabani et al. 1988] Chakravarthy et al. 1994a] Chakravarthy et al. 1994b] Haritsa et al. 1991] Pang et al. 1992] Purimetla et al. 1994] Ulusoy Belford 1993] 2.3.2 Concurrency Control Concurrency control in traditional database systems aims to ....

....by applying time critical scheduling methods to observe timing constraints of transactions. The results of the substantial amount of research devoted to the development and performance evaluation of concurrency control protocols have appeared in the following papers: Abbott GarciaMolina 1988] [Abbott Garcia Molina 1989], Abbott Garcia Molina 1992] Agrawal et al. 1992] Bestavros 1993] Bestavros and Braoudakis 1994] Biyabani et al. 1988] Buchmann et al. 1989] Chakravarthy et al. 1994a] Chakravarthy et al. 1994b] Chen Lin 1990] David et al. 1994] DiPippo Wolfe 1993] Goyal et al. 1995] ....

R. Abbott, H. Garcia-Molina `Scheduling Real-Time Transactions with Disk Resident Data', Proceedings of the 15th International Conference on Very Large Data Bases, 1989, pp.385-396.

....used to schedule the CPU in an RTDBS, priority based algorithms for concurrency control, disk scheduling, admission control, and memory management have to be developed for such a system. Over the past five years, a number of studies have investigated the problems of real time concurrency control [Abbo88a, Abbo88b, Abbo89, Hari90a, Hari90b, Hari91, Hari92, Hong93, Huan89, Huan91, Kim91] and disk scheduling [Abbo89, Abbo90, Care89, Chen91, Kim91] However, to the best of our knowledge, no work has met the admission control and memory management challenges that arise in processing queries with deadlines. These are the challenges addressed in this dissertation. 3 1.1. Real Time ....

....control, and memory management have to be developed for such a system. Over the past five years, a number of studies have investigated the problems of real time concurrency control [Abbo88a, Abbo88b, Abbo89, Hari90a, Hari90b, Hari91, Hari92, Hong93, Huan89, Huan91, Kim91] and disk scheduling [Abbo89, Abbo90, Care89, Chen91, Kim91]. However, to the best of our knowledge, no work has met the admission control and memory management challenges that arise in processing queries with deadlines. These are the challenges addressed in this dissertation. 3 1.1. Real Time Database System Architecture Jobs in a real time application ....

[Article contains additional citation context not shown here]

R. Abbott, H. Garcia-Molina, "Scheduling Real-Time Transactions with Disk Resident Data", Proc. of the 15th Int. Conf. on Very Large Data Bases, August 1989.

....of a task is defined to be the difference between its completion time and its deadline. The mean lateness for a system of tasks is the average time by which transactions miss their deadlines. The slack time of a task is an estimate of how long we can delay the task and still meet its deadline [AGM89] Three different kinds of deadlines can be identified soft, firm and hard[HCL90b] They are explained in the following paragraphs. If a task has soft deadline, the value function is often composed of two piecewise monotonically decreasing polynomials meeting at t c . The slope of the ....

....may be prohibitive. In that case, one may have to find more fine grained atomic units. As to soft and firm deadline transactions, the principle of priority inheritance has been evaluated with the underlying consistency criterion being serializability. Simulation has revealed some positive results[AGM89] We expect further improvement can be obtained if setwise serializability is used. 3.2.2.5 Predicate Correctness Korth and Speegle[KS88] proposed a criterion which combines three features to extend the serializability criterion. A schedule is said to be predicate correct when it supports ....

[Article contains additional citation context not shown here]

Robert Abbott and Hector Garcia-Molina. Scheduling real-time transactions with disk resident data. In Proc. 15th Intl Conf on Very Large Data Bases, August 1989. Shu & Young: Real-Time Concurrency Control REFERENCES 37

No context found.

R. K. Abbott and H. Garcia-Molina. Scheduling realtime transactions with disk resident data. In Proceedings of Very Large Database Conference, pages 385-- 396, 1989.

No context found.

R. K. Abbott and H. Garcia-Molina. Scheduling realtime transactions with disk resident data. In Proceedings of Very Large Database Conference, pages 385-- 396, 1989.

No context found.

R. Abbott and H. Garcia-Molina. Scheduling Real-Time Transactions with Disk Resident Data. In Very Large Data Bases, 1989.

No context found.

Abbott R., Garcia-Molina H. `Scheduling Real-Time Transactions with Disk Resident Data.' Proc. 15th Int. Conf. on Very Large Data Bases, 1989, pp.385-396.

No context found.

R. Abbott and H. Garcia-Molina. Scheduling real-time transactions with disk resident data. In Proceedings 15th International Conference on Very Large Data Bases, Los Angeles, 1989.

No context found.

R. Abbott, H. Garcia-Molina `Scheduling Real-Time Transactions with Disk Resident Data', 15th International Conference on Very Large Data Bases, pp.385-396, 1989.

No context found.

Abbott, R. and H. Garcia-Molina. (1989). "Scheduling Real-Time Transactions with Disk Resident Data." In Proc. of Conf. on Very Large Databases.

No context found.

Abbott, R. and Garcia-Molina, H. Scheduling Real-Time Transactions with Disk Resident Data. Science, Princeton, NJ, 1989.

No context found.

R. Abbott and H. Garcia-Molina, "Scheduling Real-Time Transactions with Disk Resident Data," Proceedings of the 15th VLDB Conference, 1989.

No context found.

R. Abbott and H. Garcia-Molina, "Scheduling Real-Time Transactions with Disk Resident Data," Proceedings of the 15th VLDB Conference, 1989.

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