i have seen discussions here and there about this similar issue, but i wanted to ask it again...
we are seeing this following error in both 10.2.2 environments as well as our newer 10.5.1 environments:
ExecuteError: Data access failure, layer = Image, error string = Could not access data for layer or table Image during drawingÂ§TilesWorker: Data access failure, layer = Image, error string = Could not access data for layer or table Image during drawingÂ§TilesWorker: Data access failure, layer = Image, error string = Could not access data for layer or table Image during drawing
we see this when using our python scripts to update our map services. we have two maps: one with just vector data and one with vector and raster data. our scripts get through the vector only map just fine but when its time to update the map with the raster data we get the above error.
it seems like a fairly straight forward error...well kind of. it would appear to be a data access issue, but the odd thing is we haven't changed the data or where its located or the permissions of said location. so i'm a bit perplexed on this issue, and any help would be greatly appreciated.
thanks a bunch!
we are seeing this issue again.
our python script is running the ManageMapServerCacheTiles GP Tool to rebuild new tiles based upon a delta check that we do to see what data is new.
Are you updating your production cache? If so, are you using exploded or compact cache?
If you have a dev and/or test environment, have you tried running the python script on that cache and then just copy the updated tiles from the dev/test location to the production location?
1) this is a production environment
a. this process passed in our own internal environment and pre-Production environment.
b. same script, same data
2) we have always used compact cache
3) we build the cache manually first within ArcMAP using GP Tools
a. once this process is completed, we run the script to only update tiles where data has changed
How often does this script run (e.g. daily, weekly, monthly, other)?
Does the script run at a time when no one would be accessing the cache in either a desktop or web application?
The script is set to run once a week and will run late at night unless I force the script to run. Also, this process all happens on a machine we call the maintenance server. no one is accessing anything on this machine, I am the only one running the process.
What kind of raster data are you using that gets updated on a weekly basis? I ask because my org has aerial photos that do not change plus we use an Image Server role of ArcGIS Server that is sourced from raster mosaic datasets (RMDs). Are you using RMDs for your imagery?
Have you ever thought of just removing the raster data from the mxd that is cached and just supply the RMD from an image service? With this architectural change, you remove the raster liability from your python script as it is the raster data that seems to be the root cause of the error. Image services are very efficient at supplying raster data and are equal if not faster than a cache from my experience.