A Java System That Combines Optimism and Pessimism in Updating Replicas

In a distributed system some or all of the data items are replicated and stored at separate sites. This increases the availability of these items and it is then possible to complete transactions faster than in a single site system. However, the concurrent processing of transactions at separate sites can generate inconsistencies in the stored information. Several replica control techniques have been proposed to deal with the processing of transactions . Some of these techniques include certain consistency checks before processing the transaction, i.e. they are pessimistic in nature. Other techniques process the transaction and do the consistency checking afterwards, i.e. they are optimistic. The system discussed here is an attempt to show how the increased processing capabilities provided by (i) multi-threaded object-oriented programming, and by (ii) powerful computing nodes in the distributed system can be used to combine optimism and pessimism into one system. The system allows a transaction to be processed immediately (optimistically) at any individual replica as long as the transaction satisfies a cost bound. All transactions are also processed in a concurrent pessimistic manner. Only the changes made to the replicas via pessimistic processing are made permanent.