Amazon SimpleDB is on its way and it may just change how you use databases with your web applications (or even desktop apps). SimpleDB is a web service that leverages Amazon S3 and Amazon EC2 to store, process and query datasets. Currently most if not all of us use a RDBMS such as SQL Server, Oracle, MySQL or PostgreSQL to store our data, but that requires hardware and most of the time a DBA to administer database.
Can you imagine a spatial component to Amazon SimpleDB and how you’d integrate it with your workflows? My spine is tingling just thinking about the possibilities.
You can learn more about the details of SimpleDB here.
Update: The more I think about this, the more I realize how disruptive SimpleDB will be. It was designed to be used with web applications and will be able to scale with them easily. You just can’t do that on your own. I was writing about AWS back in June and how its ability to scale could help users provide services that only the largest companies can afford.
7 responses so far ↓
1
Kristian
// Dec 14, 2007 at 3:56 pm
Amazon can haz your data? Kthxbai.
2
Jason Birch
// Dec 14, 2007 at 6:44 pm
Yes, I believe that Google has really dropped the ball on not developing Google Base to its potential. It has some simple spatial proximity query ability, but its query language is severely limited, and doesn’t appear to have seen much attention recently.
Looks like SimpleDB may be set to take the lead. Of course, FREE is compelling, so Google could still catch up
3
Matthew Snape
// Dec 15, 2007 at 4:04 am
This does look very useful. Presumably you could store and retrieve WKT strings on it. Using openlayers you could create a very basic vector GIS without needing any server software. Although indexing such data could be challenging!
4
Reid Watkins
// Dec 16, 2007 at 9:55 pm
What can we do to encourage Amazon to start spatially enabling SimpleDB now during its infancy rather than as an after thought some number of years from now?
5
Matthew Snape
// Dec 17, 2007 at 5:44 am
@Reid Watkins.
SimpleDB is not a traditional database. For example query time limits of 5 seconds exist. Even simple identify queries can take longer than that. Merely porting current GIS tecniques to a new platform is inappropriate and counterproductive. Especally when those tecniques often have less than perfect results.
6
atlasalive
// Dec 17, 2007 at 8:00 am
I’m a big fan of AWS and was pretty excited about the prospects, but I think its going to take a while before it gets all worked out. For instance, here’s a definition of their approach to data consistency from their developer’s glossary. Not exactly transaction-safe!
eventual consistency
Amazon SimpleDB keeps multiple copies of each domain. When data is written or updated (using PutAttributes, DeleteAttributes, CreateDomain or DeleteDomain) and “Success” is returned, all copies of the data are updated. However, it takes time for the data to propogate to all storage locations. The data will eventually be consistent, but an immediate read might not show the change.
Consistency is usually reached within seconds, but a high system load or network partition might increase this time. Performing a read after a short period of time should return the updated data.
7
Tim Maddle
// Dec 19, 2007 at 6:16 pm
I think this has a lot of interesting possiblities for webapps. One great addition would be a “row number between x and y” where clause.
@Matthew Snape,
If done properly, simple identify queries should be blazingly fast, certainly doable within five seconds. Using indexed shapefiles with MapServer, large spatial queries (i.e. Select By Polygon), can be done in very short order - well under 5 seconds. If whatever logic Mapserver/Mapscript uses for working with those indexed shapefiles can be ported to a project like AWS, I think you could have a very viable online spatial DB. However, I doubt the spatial aspects are at the top of Amazon’s new feature list.
Leave a Comment