Take the 2-minute tour ×
Geographic Information Systems Stack Exchange is a question and answer site for cartographers, geographers and GIS professionals. It's 100% free, no registration required.

I am new to ESRI GeoProcessing (GP) services so please excuse the naive question. I have an FLEX application running off of ArcGIS Server. I need to push data from the ArcGIS Server to an analytic server that calculates all sorts of interesting metrics using R that I cannot do within ArcGIS.

So my question is this: Can a GP accept user inputs (e.g. run on these 5 points) and then push a foreign key to another server for further processing ? The other server would do its magic and then send the results (and foreign key) back to ArcGIS server which displays to the user.

Can someone point me to some resources on the web I can read up on this ?

Thanks !

-Don

share|improve this question
add comment

1 Answer 1

You are not going to find any documentation or guidance per'se for the workflow you have described. It is a very realistic workflow though while you will probably need to use a two or step process to perform what you seek.

I see what you are seeking happening thus:

Step 1. ArcGIS request to perform ESRI specific functions and write data to common database and pass keyID to R App for next Step
Step 2.R Server performs analysis and writes data to common database and sends keyID to ArcGIS Service.
Step 3. ArcGIS Server reads keyID and Renders appropriate data back to user from common DB.

We do some common things in several of our apps that have intermediary steps that require a hand-off between processes.

share|improve this answer
    
I had a bad feeling about this - I was hoping to contain most of the communication in the web environment and push it from ESRI to the R Server without writing to a common database. I was kinda planning on using the File Geodatabase API to push from ESRI to R but some of our programmers were hoping to bypass that. Do you have some running examples I could see in action ? –  user14706 Jan 29 '13 at 22:35
    
You can possibly do that; but since you are passing between different servers/services you run into the issue of latency and performance. If you run everything inside of one machine, with the local python instance doing your interface to R then you might be able to shorten the loop up. –  D.E.Wright Jan 29 '13 at 22:39
1  
Oh man, if I give this advice, they will think you are a plant - this is EXACTLY what I was proposing. Move out of R into PySAL and run it that way (shortening the loop). DE I will try to email you off the list. –  user14706 Jan 29 '13 at 22:48
add comment

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.