1

I need to process a large dataset. Aside from processing locally on my computer, can I install, for instance, arcgis10 on a cloud service like amazon. I don't have esri server, only arcgis10.

I'm not interested in storage or distribution, just processing.

I have ample bandwidth so upload/download isn't an issue.

Might seem like a silly question, but thought it was worth asking.

Improve this question

1 Answer 1

Reset to default
1

ArcGIS Server 10 (SP4) is on Amazon EC2 (Cloud Services)

http://aws.amazon.com/solutions/global-solution-providers/esri/

http://help.arcgis.com/en/arcgisserver/10.0/help/arcgis_server_on_amazon_ec2/index.html#/What_is_ArcGIS_Server_on_Amazon_EC2/00rq00000002000000/

It has been available since July 2010

http://www.esri.com/news/releases/10_3qtr/server10-amazon.html

More info (PDF) http://www.esri.com/library/brochures/pdfs/arcgis-server-functionality-matrix.pdf

Free 60 Day Trial Evaluation of ArcGIS Server in Amazon EC2 You can request a 60-day evaluation license for Amazon EC2 running ArcGIS Server. Contact your local ESRI representative for details.

Improve this answer
6
  • The cost of server is prohibitive. But your last sentence may provide a short term solution.... Thanks for pointing that out.
    – Thad
    Commented May 23, 2012 at 15:13
  • I wonder if there are any consulting firms that will let you run a process for a fee...
    – Thad
    Commented May 23, 2012 at 15:15
  • May be it is possible to install ArcGIS Desktop on any of your nonproduction servers? Especially if you have floating license. Or any rarely used workstation PC probably? Then you can leave it processing your data during weekend for example... I am trying just to understand the problem :) Can you clarify a little?
    – Alex Markov
    Commented May 23, 2012 at 15:22
  • Sure. I'm not a newbie to processing large data, I can get it done, I was interested if there was a better (faster/more stable) way. I'm running a calculation on a raster dtm. The main issue I have with processing it over a weeks time is the time cost of making mistake or encountering a computer problem...
    – Thad
    Commented May 23, 2012 at 15:32
  • Have you considered to subdivide your data into several smaller parts? I have recently performed some processing with vector data and have discovered that processing of 1M features took ~15 min and the same processing of 500k features took ~3 min. Then I came to a solution of writing script for partitioning data before processing and collecting results after it (because actually I had to process 50M of data). Of course, I don't know if such solution is possible in your case... Anyway just some thoughts.
    – Alex Markov
    Commented May 23, 2012 at 19:34

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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