Dağıtılmış veri tabanı sistemlerinde eşzamanlılık kontrolü
Dağıtılmış veri tabanı sistemlerinde eşzamanlılık kontrolü
Dosyalar
Tarih
1995
Yazarlar
Yılmaz, Güray
Süreli Yayın başlığı
Süreli Yayın ISSN
Cilt Başlığı
Yayınevi
Fen Bilimleri Enstitüsü
Özet
The goal of this study is to investigate the implementation issues of a centralized concurrency control algoithm based on two-phase locking for DDBMS on a single processer system, namely a personel computer, running under the DOS operating system. According to this algorithm, a single cite is delegated lock management responsibility. This means that only one site has a lock manager in distributed system. Before accessing data at any site, appropri ate locks must be obtained from the central lock manager. Altough there are several proposed locking based concurrency control algorithms, the basic one of them is choosen for this simulation. Because this algorithm is easily extend, for example, primary copy and distributed concurrency control algorithms. In primary copy two phase locking concurrency control algorithm, the only changes is that since every site has a lock manager, the primary copy locations have to be determined for each data item prior to sendidng a lock or unlock request to lock manager at a site. In a distributed two phase locking algorithm, if the database is not replicated, it degenerates into the primary copy two phase locking algorithm. XI "lock" with each lock unit, this lock is set by a transaction before it is accessed and is reset at the end of its use. Obviously, a lock unit cannot be accessed by an operation if it is already locked. Thus a lock request by a transaction is granted only if the associated lock is not being held by any other transaction. Unlike the locking-based algorithms, timestamp- based concurrency control algorithms do not attempt to maintain serializability by mutual exclusion. Instead, they select, a priory a serialization order and execute transactions accordingly. To establish this ordering, the transaction manager assigns each transaction a unique timestamp. Deadlock is the one of the most important problem in the locking based concurrency control algorithms. Therefore, the distributed DBMS requires special procedures to handle this problem. Any locking-based concurrency control algorithm may result in deadlocks, since there is mutual exclusion of access to shared resources (data) and transactions may wait on locks. A deadlock can occur because transactions wait for one another. The deadlock problem in DDBSs is similar in nature to that encountered in operating systems. The competi tion among users for access to a set of resources (data, in this case) can result in deadlock if the synchronization mechanism is based on locking. The well known alternatives of prevention, avoidance and detection/recovery also apply to DDBSs. The goal of this study is to investigate the implementation issues of a centralized concurrency control algoithm based on two-phase locking for DDBMS on a single processer system, namely a personel computer, running under the DOS operating system. According to this algorithm, a single cite is delegated lock management responsibility. This means that only one site has a lock manager in distributed system. Before accessing data at any site, appropri ate locks must be obtained from the central lock manager. Altough there are several proposed locking based concurrency control algorithms, the basic one of them is choosen for this simulation. Because this algorithm is easily extend, for example, primary copy and distributed concurrency control algorithms. In primary copy two phase locking concurrency control algorithm, the only changes is that since every site has a lock manager, the primary copy locations have to be determined for each data item prior to sendidng a lock or unlock request to lock manager at a site. In a distributed two phase locking algorithm, if the database is not replicated, it degenerates into the primary copy two phase locking algorithm. XI "lock" with each lock unit, this lock is set by a transaction before it is accessed and is reset at the end of its use. Obviously, a lock unit cannot be accessed by an operation if it is already locked. Thus a lock request by a transaction is granted only if the associated lock is not being held by any other transaction. Unlike the locking-based algorithms, timestamp- based concurrency control algorithms do not attempt to maintain serializability by mutual exclusion. Instead, they select, a priory a serialization order and execute transactions accordingly. To establish this ordering, the transaction manager assigns each transaction a unique timestamp. Deadlock is the one of the most important problem in the locking based concurrency control algorithms. Therefore, the distributed DBMS requires special procedures to handle this problem. Any locking-based concurrency control algorithm may result in deadlocks, since there is mutual exclusion of access to shared resources (data) and transactions may wait on locks. A deadlock can occur because transactions wait for one another. The deadlock problem in DDBSs is similar in nature to that encountered in operating systems. The competi tion among users for access to a set of resources (data, in this case) can result in deadlock if the synchronization mechanism is based on locking. The well known alternatives of prevention, avoidance and detection/recovery also apply to DDBSs. The goal of this study is to investigate the implementation issues of a centralized concurrency control algoithm based on two-phase locking for DDBMS on a single processer system, namely a personel computer, running under the DOS operating system. According to this algorithm, a single cite is delegated lock management responsibility. This means that only one site has a lock manager in distributed system. Before accessing data at any site, appropri ate locks must be obtained from the central lock manager. Altough there are several proposed locking based concurrency control algorithms, the basic one of them is choosen for this simulation. Because this algorithm is easily extend, for example, primary copy and distributed concurrency control algorithms. In primary copy two phase locking concurrency control algorithm, the only changes is that since every site has a lock manager, the primary copy locations have to be determined for each data item prior to sendidng a lock or unlock request to lock manager at a site. In a distributed two phase locking algorithm, if the database is not replicated, it degenerates into the primary copy two phase locking algorithm.
Açıklama
Tez (Yüksek Lisans) -- İstanbul Teknik Üniversitesi, Fen Bilimleri Enstitüsü, 1995
Anahtar kelimeler
Bilgisayar ve Kontrol,
Dağıtık veri tabanı sistemleri,
Eş zamanlı programlama,
Veri tabanı sistemleri,
Computer Science and Control,
Distributed database systems,
Simultaneous programming,
Database systems