The Experts below are selected from a list of 321 Experts worldwide ranked by ideXlab platform
Haengrae Cho - One of the best experts on this subject based on the ideXlab platform.
-
Database Recovery using incomplete page versions in a multisystem data sharing environment
Information Processing Letters, 2002Co-Authors: Haengrae ChoAbstract:Coupling multiple computing nodes for transaction processing for the sake of capacity and availability has received considerable interest. A multisystem data sharing environment (MDSE) is one of the representative architecture of the coupling [2,4]. In the MDSE, all coupled nodes are connected via a high-speed network and share a common Database at the disk level. Each node in the MDSE has its own buffer pool and caches Database pages in the buffer. Caching may substantially reduce the number of disk I/O operations by utilizing the locality of reference. However, since a particular page may be simultaneously cached in different nodes, modification of the page in any buffer invalidates copies of that page in other nodes. This necessitates the use of a cache coherency scheme so that the nodes always see the most recent version of Database pages [3,6]. The subject of this paper is about Database Recovery that has to be performed when one or more nodes are failed. The complexity of Recovery schemes in the MDSE is heavily dependent on the dirty page
-
prototyping phlox a high performance transaction processing system on a workstation cluster with shared disks
IEEE Computer Society Workshop on Future Trends of Distributed Computing Systems, 2001Co-Authors: Raecha Hwang, Haengrae ChoAbstract:Coupling multiple workstations for high performance transaction processing has become increasingly attractive for reasons of capacity, cost, and availability. PHLOX is a prototypical transaction processing system running on the workstation cluster with shared disks. PHLOX supports a message oriented file system for Database sharing, global cache coherency and concurrency control, affinity-based transaction routing, and sophisticated Database Recovery protocol. The paper describes the current prototype of PHLOX and discusses some of the major design decisions that went into its construction. The lessons learned from this prototype and its predecessors are also presented.
-
FTDCS - Prototyping PHLOX, a high performance transaction processing system on a workstation cluster with shared disks
Proceedings Eighth IEEE Workshop on Future Trends of Distributed Computing Systems. FTDCS 2001, 1Co-Authors: Kyungoh Ohn, Raechan Hwang, Haengrae ChoAbstract:Coupling multiple workstations for high performance transaction processing has become increasingly attractive for reasons of capacity, cost, and availability. PHLOX is a prototypical transaction processing system running on the workstation cluster with shared disks. PHLOX supports a message oriented file system for Database sharing, global cache coherency and concurrency control, affinity-based transaction routing, and sophisticated Database Recovery protocol. The paper describes the current prototype of PHLOX and discusses some of the major design decisions that went into its construction. The lessons learned from this prototype and its predecessors are also presented.
Vijay Kumar - One of the best experts on this subject based on the ideXlab platform.
-
Recovery in Informix
Database Recovery, 1998Co-Authors: Vijay Kumar, Sang H. SonAbstract:Informix is a large scale relational Database management system. It is based on client/server paradigm and its server is referred to as “OnLine Dynamic Server” or simply “OnLine” [6]. One of the major responsibilities of OnLine is to handle Database Recovery. Similar to other commercial systems, Informix uses WAL scheme for managing commitment of concurrent transactions and Database Recovery. It uses logical and physical logs for managing these activities. We begin with the log and its management.
-
Formalization of Recovery
Database Recovery, 1998Co-Authors: Vijay Kumar, Sang H. SonAbstract:Database Recovery guarantees the consistency of the Database in the event of transaction, system, and media failure, any of which may happen for a variety of reasons. Robust Recovery is, therefore, an essential feature of any DBMS. However, in spite of its important role in the functionality of DBMS and vast experience in building Recovery systems, the process of Recovery has not been treated on a conceptual level. A careful review of the literature on Recovery reveals that there is a semantic gap between high level requirements such as the all-or-nothing property, and recoverability and how these requirements are implemented, in terms of buffers and their policies, volatile and persistent storage, shadowing, etc. The advent of new transaction models for dealing with nontraditional applications makes the Recovery even more important, and, therefore, elevating Recovery from a process to a concept is desirable.
-
Recovery mechanisms in Database systems
1997Co-Authors: Vijay Kumar, Meichun HsuAbstract:From the Publisher: This is the first book to bring together all you need to know about Database Recovery - both theory and practice. It covers not just Recovery mechanisms in today's relational Databases, but also the critically important new techniques for main memory Databases, mobile computing, and enterprise workflow systems. Recovery Mechanisms in Database Systems will be an essential resource for all Database designers, architects, decision-makers, and advanced students in the field.
-
DEXA - Recovery in Main Memory Database Systems
Lecture Notes in Computer Science, 1996Co-Authors: Vijay KumarAbstract:In this paper we present a Recovery mechanism for main memory Database, which does not treat volatile RAM as a buffer and uses a limited size non-volatile RAM for efficient logging and archiving. It allows Database Recovery in parallel to the execution of normal transactions, it combines Undo and Redo into a single operation in many situations and eliminates the need for checkpointing. Under partial Recovery our algorithm manages to identify and recovers only dirty data for bringing the entire Database into a consistent state.
-
SPDP - Main memory Database Recovery algorithms and their performance
Proceedings of the Second IEEE Symposium on Parallel and Distributed Processing 1990, 1Co-Authors: Vijay Kumar, A. BurgerAbstract:In main memory systems Recovery becomes more complex mainly due to the volatility of main memory and the elimination of the separation of data storage (disk) and data processing locations (main memory). In a disk based Database system, it is the disk copy that represents the stable Database restored after a system failure; but in MDBSs, the main memory copy of the Database must be recovered. The paper discusses the Recovery issues in MDBSs and presents a somewhat detailed performance study of two Recovery algorithms, one using a shadow approach and the other update-in-place. >
A. Burger - One of the best experts on this subject based on the ideXlab platform.
-
Performance measurement of main memory Database Recovery algorithms based on update-in-place and shadow approaches
IEEE Transactions on Knowledge and Data Engineering, 1992Co-Authors: V. Kumar, A. BurgerAbstract:The authors present a performance study of two main memory Database Recovery algorithms, one using a shadow approach and the other update-in-place. The results show that in main memory Databases, shadow approach performs better than update-in-place. The authors introduce minor improvements to the shadow scheme and show that the modified algorithm does show further performance improvement.
-
SPDP - Main memory Database Recovery algorithms and their performance
Proceedings of the Second IEEE Symposium on Parallel and Distributed Processing 1990, 1Co-Authors: Vijay Kumar, A. BurgerAbstract:In main memory systems Recovery becomes more complex mainly due to the volatility of main memory and the elimination of the separation of data storage (disk) and data processing locations (main memory). In a disk based Database system, it is the disk copy that represents the stable Database restored after a system failure; but in MDBSs, the main memory copy of the Database must be recovered. The paper discusses the Recovery issues in MDBSs and presents a somewhat detailed performance study of two Recovery algorithms, one using a shadow approach and the other update-in-place. >
-
ICDE - Performance measurement of some main memory Database Recovery algorithms
[1991] Proceedings. Seventh International Conference on Data Engineering, 1Co-Authors: Vijay Kumar, A. BurgerAbstract:The performance of two main-memory Database system (MDBS) algorithms is studied. M.H. Eich's (Proc. of the Fifth Int. Workshop on Database Machines., Oct. 1987) and T.J. Lehman's (Ph. D. Thesis, Univ. of Wisconsin-Madison, Aug 1986) algorithms are selected for this study. The investigation indicates that the shadow approach, as used by Eich, has some advantages over the update in-place strategy of Lehman. The shadow approach has faster response for normal transactions even though transaction commit is slower compared to update in-place approach. It is concluded that irrespective of the Recovery algorithm an efficient load balancing plays an important part in the performance of the system. It is suggested that processors should be allocated to any kind of activity on a demand basis. It is discovered that individual Recovery of lockable units as used in Lehman's algorithm is a better choice since it increases the system availability after a system failure. It also allows background Recovery to go in parallel to normal transaction processing. >
Kian-lee Tan - One of the best experts on this subject based on the ideXlab platform.
-
SIGMOD Conference - Fast Failure Recovery for Main-Memory DBMSs on Multicores
Proceedings of the 2017 ACM International Conference on Management of Data, 2017Co-Authors: Wentian Guo, Chee-yong Chan, Kian-lee TanAbstract:Main-memory Database management systems (DBMS) can achieve excellent performance when processing massive volume of on-line transactions on modern multi-core machines. But existing durability schemes, namely, tuple-level and transaction-level logging-and-Recovery mechanisms, either degrade the performance of transaction processing or slow down the process of failure Recovery. In this paper, we show that, by exploiting application semantics, it is possible to achieve speedy failure Recovery without introducing any costly logging overhead to the execution of concurrent transactions. We propose PACMAN, a parallel Database Recovery mechanism that is specifically designed for lightweight, coarse-grained transaction-level logging. PACMAN leverages a combination of static and dynamic analyses to parallelize the log Recovery: at compile time, PACMAN decomposes stored procedures by carefully analyzing dependencies within and across programs; at Recovery time, PACMAN exploits the availability of the runtime parameter values to attain an execution schedule with a high degree of parallelism. As such, Recovery performance is remarkably increased. We evaluated PACMAN in a fully-fledged main-memory DBMS running on a 40-core machine. Compared to several state-of-the-art Database Recovery mechanisms, can significantly reduce Recovery time without compromising the efficiency of transaction processing.
-
Parallel Database Recovery for Multicore Main-Memory Databases.
arXiv: Databases, 2016Co-Authors: Wentian Guo, Chee-yong Chan, Kian-lee TanAbstract:Main-memory Database systems for multicore servers can achieve excellent performance when processing massive volume of OLTP workloads. But crash-resilience mechanism, or namely loggingand-replay, can either degrade the performance of transaction processing or slow down the process of failure Recovery. In this paper, we show that, by exploiting application semantics, it is possible to achieve speedy failure Recovery without introducing any costly logging overhead to the execution of online transactions. We propose TARS, a parallel log-Recovery mechanism that is specifically designed for lightweight, coarse-grained command-logging approach. TARS leverages a combination of static and dynamic analyses to parallelize the log replay: at compile time, TARS decomposes stored procedures by carefully analyzing dependencies within and across programs; at Recovery time, a re-execution schedule with a high degree of parallelism is attained through lightweight one-pass scans over transaction parameter values. As such, Recovery latency is remarkably minimized. We evaluated TARS in a main-memory Database prototype running on a 40-core server. Compared to the state-of-the-art mechanisms, TARS yields significantly higher Recovery rate without compromising the efficiency of transaction processing.
-
Fast Failure Recovery for Main-Memory DBMSs on Multicores
arXiv: Databases, 2016Co-Authors: Wentian Guo, Chee-yong Chan, Kian-lee TanAbstract:Main-memory Database management systems (DBMS) can achieve excellent performance when processing massive volume of on-line transactions on modern multi-core machines. But existing durability schemes, namely, tuple-level and transaction-level logging-and-Recovery mechanisms, either degrade the performance of transaction processing or slow down the process of failure Recovery. In this paper, we show that, by exploiting application semantics, it is possible to achieve speedy failure Recovery without introducing any costly logging overhead to the execution of concurrent transactions. We propose PACMAN, a parallel Database Recovery mechanism that is specifically designed for lightweight, coarse-grained transaction-level logging. PACMAN leverages a combination of static and dynamic analyses to parallelize the log Recovery: at compile time, PACMAN decomposes stored procedures by carefully analyzing dependencies within and across programs; at Recovery time, PACMAN exploits the availability of the runtime parameter values to attain an execution schedule with a high degree of parallelism. As such, Recovery performance is remarkably increased. We evaluated PACMAN in a fully-fledged main-memory DBMS running on a 40-core machine. Compared to several state-of-the-art Database Recovery mechanisms, PACMAN can significantly reduce Recovery time without compromising the efficiency of transaction processing.
Francesc D. Muñoz-escoí - One of the best experts on this subject based on the ideXlab platform.
-
ISPDC - Optimizing Certification-Based Database Recovery
Sixth International Symposium on Parallel and Distributed Computing (ISPDC'07), 2007Co-Authors: J. Pla-civera, M. I. Ruiz-fuertes, Luis H. García-muñoz, Francesc D. Muñoz-escoíAbstract:Certification-based Database replication protocols are a good basis to develop replica Recovery when they provide the snapshot isolation level. For such isolation level, no readset needs to be transferred between replicas nor checked in the certification phase. Additionally, these protocols need to maintain a historic list of writesets that is used for certifying the transactions that arrive to the commit phase. Such historic list can be used to transfer the missed state of a recovering replica. We study the performance of the basic Recovery approach - to transfer all missed writesets - and a version-based optimization - to transfer the latest version of each missed item, compacting thus the writeset list - and the results show that such optimization reduces a lot the Recovery time.
-
OTM Conferences (1) - Reviewing amnesia support in Database Recovery protocols
On the Move to Meaningful Internet Systems 2007: CoopIS DOA ODBASE GADA and IS, 2007Co-Authors: Rubén De Juan-marín, José Enrique Armendáriz-iñigo, Luis H. García-muñoz, Francesc D. Muñoz-escoíAbstract:Replication is used for providing highly available and fault-tolerant information systems, which are constructed on top of replication and Recovery protocols. An important aspect when designing these systems is the failure model assumed. Replicated Databases literature last trends consist in adopting the crash-Recovery with partial amnesia failure model because in most cases it shortens the Recovery times. But, despite the large use of such failure model we consider that most of these works do not handle accurately the amnesia phenomenon. Therefore, in this paper we survey some works, analyzing their amnesia support.
-
ISCIS - A protocol for reconciling Recovery and high-availability in replicated Databases
Computer and Information Sciences – ISCIS 2006, 2006Co-Authors: José Enrique Armendáriz-iñigo, Francesc D. Muñoz-escoí, H. Decker, J. R. Juárez-rodríguez, J. R. González De MendívilAbstract:We describe a Recovery protocol which boosts availability, fault tolerance and performance by enabling failed network nodes to resume an active role immediately after they start recovering. The protocol is designed to work in tandem with middleware-based eager update-everywhere strategies and related group communication systems. The latter provide view synchrony, i.e., knowledge about currently reachable nodes and about the status of messages delivered by faulty and alive nodes. That enables a fast replay of missed updates which defines dynamic Database Recovery partition. Thus, speeding up the Recovery of failed nodes which, together with the rest of the network, may seamlessly continue to process transactions even before their Recovery has completed. We specify the protocol in terms of the procedures executed with every message and event of interest and outline a correctness proof.
-
PDP - CLOB: communication support for efficient replicated Database Recovery
13th Euromicro Conference on Parallel Distributed and Network-Based Processing, 1Co-Authors: Francisco Castro-company, M. I. Ruiz-fuertes, Javier Esparza-peidro, Luis Irun-briz, H. Decker, Francesc D. Muñoz-escoíAbstract:Replication protocols using an eager update propagation strategy commonly need a reliable broadcast service; i.e., a broadcast primitive with atomic delivery and, in some cases, also with total order. This communication service provides some appropriate features for the Recovery tasks, although in some cases this leads to partial blocking of the replica taken as the source in the updating process. CLOB is a framework for reliable broadcast protocols that log the missed update messages in case of failure, being able to automatically resend these updates when the faulty destinations recover. This behaviour is easily configurable and allows an efficient Recovery mechanism in case of short-term failures, which can be combined with other version-based Recovery protocols in order to manage long-term outages.
-
OTM Conferences (1) - Revisiting certification-based replicated Database Recovery
On the Move to Meaningful Internet Systems 2007: CoopIS DOA ODBASE GADA and IS, 1Co-Authors: M. I. Ruiz-fuertes, J. Pla-civera, José Enrique Armendáriz-iñigo, J. R. González De Mendívil, Francesc D. Muñoz-escoíAbstract:Certification-based Database replication protocols are a good means for supporting transactions with the snapshot isolation level. Such kind of replication protocol does not demand readset propagation and allows the usage of a symmetric algorithm for terminating transactions, thus eliminating the need of a final voting phase. Recovery mechanisms especially adapted for certification-based replication protocols have not been thoroughly studied in previous works. In this paper we propose two Recovery techniques for this kind of replication protocols and analyze their performance. The first technique consists in dividing the Recovery in two stages, reducing the certification load and the amount of information to be recovered in the second stage. The second technique scans and compacts the set of items to transfer, sending only the latest version of each item. We show that these techniques can be easily combined, reducing thus the Recovery time.
