Update propagation protocols for replicated databates

Abstract

Replication is often used in many distributed systems to provide a higher level of performance, reliability and availability. Lazy replica update protocols, which propagate updates to replicas through independent transactions after the original transaction commits, have become popular with database vendors due to their superior performance characteristics. However, if lazy protocols are used indiscriminately, they can result in non-serializable executions. In this paper, we propose two new lazy update protocols that guarantee serializability but impose a much weaker requirement on data placement than earlier protocols. Further, many naturally occurring distributed systems, like distributed data warehouses, satisfy this requirement. We also extend our lazy update protocols to eliminate all requirements on data placement. The extension is a hybrid protocol that propagates as many updates as possible in a lazy fashion. We implemented our protocols on the Datablitz database system product developed at Bell Labs. We also conducted an extensive performance study which shows that our protocols outperform existing protocols over a wide range of workloads.

Pronto: High availability for standard off-the-shelf databases

“…• Locating the primary server (lines [2][3][4][5][6][7][8][9][10][11][12][13][14]. Initially, the client chooses one database server (lines 7-8), tries to start the transaction on this server (line 9), and waits for an • Executing a transactional job (lines [15][16][17][18][19][20][21][22][23][24][25].…”

“…Although the approach is particularly suited for weak consistency [47], some epidemic protocols have been augmented to ensure serializability [3]. Ensuring serializability with lazy propagation has also been achieved by restricting the placement of replicas [10,16] and building global replication graphs [11].…”

Implementing a Reliable Digital Object Archive

“…The Stanford site contacts the MIT site and discovers that MIT does not yet know about the agreement, so that all three objects are replicated to MIT. 3 Similarly, all three objects are copied to Berkeley (Figure 4b). Each of the three sites then begins a cycle of repeated site checks, connecting to the other two sites and comparing snapshots.…”

“…Much work has been done in designing algorithms for propagating data from one replicate to another [3,33]. These systems focus on systems that allow updates and deletions of objects.…”

Grid computing: the European Data Grid Project

“…Distributed database research deals with update synchronisation of single transactions on replicas [13]. In [14] Internet specific protocols and worldwide load balancing are discussed.…”