Menu
Home
People
Places
Arts
History
Plants & Animals
Science
Life & Culture
Technology
Reference.org
Distributed concurrency control
open-in-new
<p>Distributed concurrency control is the <a href="/facts/Concurrency_control/JWnmhVd5">concurrency control</a> of a system <a href="/facts/Distributed_computing/o5nyhqTF">distributed</a> over a <a href="/facts/Computer_network/3w5RM99p">computer network</a> (Bernstein et al. 1987, Weikum and Vossen 2001). </p><p>In <i><a href="/facts/Database_systems/VYcpkevb">database systems</a></i> and <i><a href="/facts/Transaction_processing/Q0WHeD9R">transaction processing</a></i> (<i>transaction management</i>) distributed concurrency control refers primarily to the concurrency control of a <a href="/facts/Distributed_database/A6Q1s4Iy">distributed database</a>. It also refers to the concurrency control in a multidatabase (and other multi-transactional object) environment (e.g., <a href="/facts/Federated_database/SSFWzg6k">federated database</a>, <a href="/facts/Grid_computing/2GQhzN5l">grid computing</a>, and <a href="/facts/Cloud_computing/guhg0FGV">cloud computing</a> environments. A major goal for distributed concurrency control is distributed <a href="/facts/Serializability/j6LY0fqE">serializability</a> (or <a href="/facts/Global_serializability/XhFwnk4y">global serializability</a> for multidatabase systems). Distributed concurrency control poses special challenges beyond centralized one, primarily due to communication and computer <a href="/facts/Latency_(engineering)/6Bjhn30H">latency</a>. It often requires special techniques, like <a href="/facts/Distributed_lock_manager/CGvJQHH9">distributed lock manager</a> over fast <a href="/facts/Computer_network/3w5RM99p">computer networks</a> with low latency, like <a href="/facts/Switched_fabric/TwfTsu3e">switched fabric</a> (e.g., <a href="/facts/InfiniBand/6CUz47wQ">InfiniBand</a>). </p><p>The most common distributed concurrency control technique is <i>strong strict two-phase locking</i> (<a href="/facts/Two_phase_locking/A1Ezfv2n">SS2PL</a>, also named <i>rigorousness</i>), which is also a common centralized concurrency control technique. SS2PL provides both the <i>serializability</i> and <i><a href="/facts/Database_transaction_schedule/9kRscku5">strictness</a></i>. Strictness, a special case of recoverability, is utilized for effective recovery from failure. For large-scale distribution and complex transactions, distributed locking's typical heavy performance penalty (due to delays, latency) can be saved by using the atomic commitment protocol, which is needed in a distributed database for (distributed) transactions' <a href="/facts/Atomicity_(database_systems)/A356GsoW">atomicity</a>. </p>