biovel-nbc

Naturalis implementations of BioVeL services.

Asynchronous API to be implemented

One endpoint to submit the different jobs, with four parameters:

• name --- required
• arguments --- required
• sessionid --- optional
• email --- optional

name differentiates the different types of work, while arguments give the info. Data are given as URL(s) within arguments, which allows multiple submission divided by ";". Within phylogenetics service set we are not using yet this multiple submission proposed by INFN. The reply of this insertjob is an xml with several tags. Example:

<Job>
	<Name>blast</Name>
	<Flag>7772590c-d7ab-4b76-98b9-aa293c6c34fe</Flag>
	<JobsID>
		<JobId>454118</JobId>
	</JobsID>
</Job>

One endpoint to collect results with one parameter, jobid, that identifies the job to retrieve. The reply is for example:

<Jobs>
	<Job>
		<Arguments>
			http://webtest.ba.infn.it/vicario/FinalFusariumDB_2.nex 
			5700.fa 10589.fbsse se selknlk noiho niooih r 
		</Arguments>
		<Comment>interactive</Comment>
		<CPUs>7</CPUs>
		<Flag>2dbbd030-2803-43c0-a21d-369a21e17f2b</Flag>
		<Id>349177</Id>
		<LastCheck>2011-11-18 12:54:01.0</LastCheck>
		<Name>MyBlasts</Name>
		<Output/>
		<Provenance/>
		<Status>free</Status>
	</Job>
</Jobs>

The workflow is looking for these paths:

1. /Jobs/Job/Status to check if is equal to done
2. /Jobs/Job/Output to collect the body that could be or an xml or url or a base64 encoded string. Generally is a url with bulk information not relevant for the taverna engine but to pass on to next webservice.
3. /Jobs/Job/StandardOutput to collect the body that could be or an xml or url or a base64 enconded string. Generally is a xml with information to be processed within the taverna engine
4. /Jobs/Job/StandardError to collect the body that is a base64 encoded string with all relevant info for debug error