Ramesh Kumar Gupta, "Commit Processing in Distributed On-Line and Real-Time Transaction Processing Systems", MSc Thesis, Supercomputer Education and Research Centre, Indian Institute of Science, 1997.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....SRTDB system is described in Chapter 5, and the results of the experiments are presented in Chapter 6. Finally, in Chapter 7, we provide an overall summary of the work presented in this thesis and identify future research avenues. Major parts of the material for this thesis has been taken from [30] and [31] Secure Real Time Distributed Databases 2.1 Introduction The secure real time transaction processing domain is a complex union of database technology, real time technology and information security. The purpose of this chapter is to review the basics of these three areas and then to ....

....abort and presumed commit protocols, and the three phase commit protocol. Then, we will describe a new protocol, called OPT, that allows the transactions to access uncommitted data in a controlled fashion in order to improve the performance of the system.Material for this chapter is taken from [30] 3.1 Recovery Protocols for Distributed Databases The recovery manager provides the semantics of atomicity and durability to the transactions. All of the effects of the committed transactions and none of the effects of the aborted transactions must be visible to the database. This must be ....

[Article contains additional citation context not shown here]

Ramesh Kumar Gupta, "Commit Processing in Distributed On-Line and Real-Time Transaction Processing Systems", MSc Thesis, Supercomputer Education and Research Centre, Indian Institute of Science, 1997.

....in distributed RTDB systems by performing an extensive set of simulation experiments. Our contributions are two fold: ffl Using a detailed model of a distributed firm deadline real time database system, we profile the real time performance of a representative suite of commit protocols studied in [Gup97] and extend the study for the one phase commit protocols.To the best of our knowledge, these are the first quantitative results for the one phase protocols in this area. ffl We reveal an interesting split personality behavior of the one phase protocols. The one phase protocols perform very well ....

....maximize the number of transactions that complete before their deadlines expire, as opposed to minimizing the average transaction response time (or maximizing the average transaction throughput) which is the primary goal in a conventional database system. Material for this chapter is taken from [Gup97]. 3.1 Real Time Framework The applications in the real time domain can be grouped into the following three categories based on how the application is impacted by the violation of the task completion deadline [HCL92] Hard Deadline Real Time Applications: In these applications, the consequences ....

[Article contains additional citation context not shown here]

Ramesh Kumar Gupta. Commit Processing in Distributed On-Line and RealTime Transaction Processing Systems. Master's thesis, Supercomputing Education and Research Centre, Indian Institute of Science, Bangalore, India, 1997.

....model described in the previous section, we conducted an extensive set of simulation experiments comparing the real time performance of the 2PC, PA, PC, 3PC and PROMPT commit protocols. In this section, we present the results of a representative set of experiments (the complete set is available in [18]) A. Experiment 1: Resource and Data Contention Our first experiment was conducted using the default settings for all model parameters (Table I) resulting in significant levels of both resource contention (RC) and data contention (DC) Here, each transaction executes in a sequential fashion ....

....features of PROMPT is its ability to combine with other optimizations. We now evaluate the performance benefits possible from such combinations. G. 1 Experiment 7: PROMPT PA and PROMPT PC When we analyzed the performance effects of adding the PA and PC optimizations to PROMPT (graphs available in [18]) we observed that just as PA and PC provided little improvement over the performance of standard 2PC, here also they provide no tangible benefits to the performance of the PROMPT protocol and for the same reasons. While PROMPT PA is very slightly better than basic PROMPT, PROMPT PC performs ....

[Article contains additional citation context not shown here]

R. Gupta, "Commit Processing in Distributed On-line and RealTime Transaction Processing Systems," M.S. thesis, SERC, Indian Institute of Science, March 1997.

....distributed commit protocols and then the precise semantics of firm deadlines in a distributed environment. 3.1 The OPT Protocol The OPT protocol [10] was designed to address the above mentioned issue of prepared data blocking. The main feature of OPT (the complete description is available in [9]) is that transactions requesting data items held by other transactions in the prepared state are allowed to access this data. That is, prepared cohorts lend uncommitted data to concurrently executing transactions in the optimistic belief that this data will eventually be committed. The ....

....in the previous section, we conducted an extensive set of simulation experiments comparing the performance of the various commit protocols presented earlier. In particular, our focus is on parallel transaction workloads the corresponding results for sequential transactions are available in [9]. The performance metric in all of our experiments is MissP ercent, which is the percentage of input transactions that the system is unable to complete before their deadlines. 4 MissPercent values in the range of 0 to 20 percent are taken to represent system performance under normal loads, ....

[Article contains additional citation context not shown here]

R. Gupta, "Commit Processing in Distributed On-Line and Real-Time Transaction Processing Systems", M.Sc.(Engg.) Thesis , SERC, Indian Institute of Science, March 1997.

....(this effect was confirmed by measuring these values in our experiments) resulting in OPT having lesser impact on the throughput even though it continues to provide better throughput than 2PC and its classical variants. The detailed results for all the sequential experiments are available in [Gup97] for illustrative purposes, we present here the results for the baseline RC DC and pure DC situations in Figures 11a and 11b, respectively. CENT DPCC 2PC 3PC PC OPT 0 2 4 6 8 10 0 5 10 15 20 25 MPL Site System Throughput Fig 11a : Sequential (RC DC) 0 2 4 6 8 10 0 10 20 30 40 50 ....

....to explore various regions of the workload space. These include workloads with different database sizes, non zero think times, etc. The relative performance of the protocols in these additional experiments remained qualitatively similar to that seen in the experiments described here (see [Gup97] for the details) 6 Lending Variations of OPT In the previous section, we quantitatively demonstrated that the OPT approach is conducive to good performance. We now move on to considering possible variations, in the design of OPT s lending mechanism, that may have the potential to increase the ....

[Article contains additional citation context not shown here]

R. Gupta, "Commit Processing in Distributed On-Line and Real-Time Transaction Processing Systems", M.Sc.(Engg.) Thesis, SERC, Indian Institute of Science, March 1997.

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