Good afternoon gang!
I have been poking around all the ESRI resources, and didn't find exactly what I needed. We are looking at using the Image Server Extension to incorporate some topographic data sets as a basemap into our customized server site. Yes, I know there are already map services from ESRI and USGS etc that accomplish this. However, one of our requirements is that the SR of our basemap cannot be WGS 84,among other req's like owning our data etc. so there goes the cached service from ESRI
Currently, we have 5 Raster Catalogs each containing a specific geographic area of custom topo map images. Soon we will have 5 raster catalogs which contain imagery in this same geographic extent as the topos. These will both grow in data over the coming years
Is it possible or recommended to store our data this way if we want to serve as a single image service? I ask because I have not been able to figure out the best way to create a Referenced Mosaic using more than one Raster Catalog, hence why we only have a single image service for each Raster Catalog or ref mosaic. If our dataset will eventually grow to TB sizes, I would assume a single raster catalog would NOT be the way to go, but maybe I'm wrong?
What's the consensus out there? How should we attack this?
I apologize for the VERY long delay since you posted this question. I'll keep my response very brief but if you still need advice and information, please contact me.
We are building documentation focused on Best Practices regarding managing and serving imagery (including lidar and elevation data). The starting point is this landing page http://esriurl.com/6550 which points you to detailed discussion in a Guidebook in ArcGIS Help, and also to an AGOL Group with downloadable tools to show examples for different types of data (elevation, preprocessed orthophotos, scanned maps, etc.).
I'm adding these questions to this thread because I believe they fall under the heading of "best practices"
Cody Benkelman re: your mention of the best practices guidebook above..... for caching from a Image Service extension service, the guidebook, help, and the tech session at the UC all mention that it is best to cache the largest scale (in my case, the source resolution) and let the other cache levels be built from that, which is opposite of the thought process I have used before. In my situation, I am building the cache from a image service based on a mosaic dataset. This dataset has many gaps (i.e. data is still being collected, and some areas will not have data (ocean))
- Should I specify area-of-interest to only those areas that I know there is data, and if so, is there a way of specifying this thru/from my mosaic dataset? OR, is the image service smart enough to handle this (that is, I don't want a large no-data area.
- Is there anything special I need to check to have the other levels work off the first (L09) that I create? Update below
- If I plan to do one level at a time (due to network resources), should I run L08 complete, then L07, etc? Update below
Update -- still have a question about specifying Area of Interest when caching, and whether it is necessary. However, the image service caching does seem to know to create the cache from L09 to L08 to L07, etc. on it's own. yea.