Can anybody explain about DS RTI Services and purpose &

[pongal]

Hi Guys,
Can anybody explain what's the main purpose of RTI Services and where it is used.

[ray.wurlod]

RTI Services allows you to expose a DataStage job as a web service with real-time response. Often the DataStage job will incorporate one or more QualityStage components, for fuzzy matching and best-record survivorship.

Imagine you have a DataStage job that processes a row of data to produce a changed row. Imagine further - though it's not necessary - that the input and output are XML.

Code: Select all

XMLReader  ----->  OtherStages  ----->  XMLWriter

To convert this for real-time operation, you surround it with a pair of RTI stages.

Code: Select all

RTIInput  ----->  XMLReader  ----->  OtherStages  ----->  XMLWriter  ----->  RTIOutput

Such a job can now receive requests from a web application (for example) that provides an XML document containing the data to be processed, and which receives XML back again.

For example, at my current site, data entry applications invoke searches (matches) and correction jobs so that misspelled names, addresses and localities can be picked up before being committed to the database.

[pongal]

Thanks for your information...
i have a doubt that shall we extract the data from pdf files by using DS RTI and load it into particular database like oracle or DB2.

[ray.wurlod]

Almost certainly NOT an appropriate use for RTI Services, there's no "real time" requirement, is there?

[mee]

Rey, thanks for the explanation of the RTI and WebService. I do have few follow on questions - I don't fully understand this yet.

1. We get many intra day falt file (and XML) drops. These files have data that are related in some form and are used to generate XML files that we would like to push to downstream application. Using your example, it would be

RTI input - File reader - Other stages - XML writer - RTI output.

Is this possible? If so, could you pls elaborate.

2. This job (that I expose as Web Service) is likely to be called by different application simultaneously. Is this possible with RTI? Will RTI have some locking related problem?

3. What extent the calling Web application is notified about the error when somethinhg goes wrong in the ETL job? For example, if my input is XML and the XML scema didn't match the input XML data, will the calling application be notified about this mismatch?

4. In case of XML input, the XML reader need to validate the XML schema before processing the data. Now suppose I have the same XML schema, but many different small xml files. If the job is invoked with different XML files, does that imply XML schema validation on every invocation or once in the beginning? The reason for my question is we have very large XML schema, but many small XML files. XML schema validation takes much longer compared to actual processing. If schema validation happens on every invocation, then it may not work for us.

Thanks for your help.