article · 2007
Using advanced transaction and workflow models in composing web services
Cited by 0Abstract and details only
Abstract
A useful feature of Web services is that new and more complex Web services can be composed of other Web services. However, in many cases composed web services are useful only if they can be processed atomically. The (semantic) atomicity of composed Web services is ensured in a similar way as with advanced transaction models which use compensating transactions to undo the effects of successfully executed subtransactions. However, a problem of such compensation is that it has two contradicting semantics: from local point of view the subtransaction is like any single transaction while from global point of view it is a subtransaction of a distributed transaction having the mutual atomicity dependencies. As a result executing compensating transaction may cause problems, e.g., the cancellation of the hotel reservation may give rise for a special charge. In order to achieve shared understanding of transactions' semantics we have developed a transaction ontology to be used in the Web. In this paper we explain how it should be used in information exchange between Web services. In addition we present our developed CWS-Transaction model (Composed Web Service Transaction model), which censures the atomicity of composed Web services but does not use compensating transactions at all. We also illustrate how CWS-transactions can be processed by a workflow engines based on BPEL4WS-code.
Cite this paper
Puustjärvi, J. (2007). Using advanced transaction and workflow models in composing web services.
Puustjärvi, Juha. “Using advanced transaction and workflow models in composing web services.” 2007, pp. 58–63.
Puustjärvi, Juha. 2007. “Using advanced transaction and workflow models in composing web services.”
Puustjärvi, J. (2007) ‘Using advanced transaction and workflow models in composing web services’.
@article{puustjarvi2007using,
title = {{Using advanced transaction and workflow models in composing web services}},
author = {Puustjärvi, Juha},
pages = {58--63},
year = {2007}
}Made from the index record. Check names, capitalisation and page numbers against the paper. Need another style? AI for Research has over 10,000.