Latest Contributions by JonathanQuinn

What are you using as the front-end to your Portal?  Are you using the web adaptor or your own reverse proxy?  If you're using the web adaptor, does it exist on the same machine or a different machine?  If it's on a different machine, is that running Windows Server 2008 as well? ... View more

HTTP/HTTPS shouldn't really matter.  If you were to capture the request from the application using Fiddler, grab the webmap_as_json parameter and then use that JSON for the same parameter back in ArcMap.  If it doesn't do anything with  your custom elements, then it won't do it as a service.  That may give you a good starting point for troubleshooting as well. ... View more

There aren't any hostname changes involved, right?  You may want to reach out to Support to see if they can dig a bit deeper. ... View more

Does it matter what browser you use? ... View more

You shouldn't need to republish the data.  From your screenshot, you do have a lot of permission related problems given the "base table definition string is invalid" error.  Are you using DB authentication or OS authentication to access your database?  Those errors would cause the data to not display at all, instead of not displaying at certain scales.  Does the layer name in the service match up with the errors in the logs about base table definition string being invalid? ... View more

Are Portal and Server federated?  Are you using a self signed certificate for the URL you're using to reach Manager?  If I'm right in assuming what the problem is, this is actually a new problem introduced by Chrome.  Chrome is forgetting the exception to the self signed certificate, so when Manager redirects to Portal to obtain authorization and returns back to Manager, the exception is forgotten and the browser doesn't proceed.  You should see insecure responses when using the dev tools in Chrome.  The workaround is to use IE or Firefox, trust the self signed certificate, or use a certificate that Chrome will trust for 6443. ... View more

I think that anonymous access to the portal means the portal website, but you can still share items publicly which is unrelated to the anonymous access setting.  If you don't want people to access the portal or items without a login, set anonymous access to disabled and only share items with the organization. ... View more

Have you taken a look at the HTTP requests that are made within the browser when going through this manually?  You can capture them using Fiddler or the dev tools within the browser you're using.  You're essentially just recreating those requests in Python. Specific to your question regarding lines 51 through 65, it starts earlier at line 39.  It seems like you have JSON that's stored in two files.  You read the contents from one file, (filein on line 43), and then replace a string, (line 46), and then save it to fileout, (line 48).  You then read the contents again on line 52, (you can actually just reuse the newdata variable from 48), and then you de-serialize the JSON into text, (line 55), and then construct the "payload" for the update request, (line 56).  Payload data has to be encoded, (line 59), and then you construct a Request object, (line 60), and then retrieve the response from the request, (line 61). You're not really getting the response of the request, so potentially, there's a problem you're not capturing.  You can just print the response after line 61 to see if the request went through.  I think you'll see {"status":"success"...} or something similar. The other thing is take a look at the requests in Fiddler or the dev tools, you may be missing a required parameter.  For example, I don't see a token passed along wiht the request, so the response is likely a 498 error, token required.  That may be handled by lines 70 and 71, but I don't think that handler is being used if you're submitting the requests yourself using urllib and urllib2.  You'll need to provide a token. ... View more

You should just be able to copy the URL of the service, (including the task name), and paste it into the URL parameter for the print task.  Are you seeing problems when doing so? ... View more

I would first make sure that the data is visible in ArcMap at the same scales you can't see it in the service.  If you can't see it, then that's an indication it's a spatial index issue.  You can try to recreate the spatial index.  You can also copy the data to create a new feature class or use the FC2FC tool, which would recreate the spatial index. If you can see the data at all scales, then it's a problem with the Server, not the data.  It's definitely not a permissions issue, as you wouldn't be able to see the data at all at any scale and the publishing process would tell you that the service couldn't come up as it couldn't find the data. Definitely look at the logs under Verbose to analyze the different operations within the export map request.  You should see data access, symbolizing, etc.  There may be errors when trying to draw the data. Also experiment with different data sources, similar to the suggestions in the first paragraph.  If it's in SDE, export it to a FGDB feature class.  If it's in an FGDB feature class, export it to a shapefile.  Then, open a new map, add the data, and publish the service. ... View more

No you can register two web adaptors with the same site and configure them with separate authentication methods.  If people external to your network will still be using Windows credentials to access the services, then you'll need to install the second web adaptor so it's available externally as well.  If the only time people will access the services will be when they're internal in your network, you can install the web adaptor on an internal web server. Once you federate, I think you can still get away with two web adaptors, but you'll need to federate using the URL that people will use to access secure services, not the URL that will be used to access unsecure services. ... View more

You're processing the data in memory, which will be faster than writing to disk.  If you want to save the results on disk, use the CopyFeatures tool and construct the path using the arcpy.env.scratch<loc> environment variable. outData = os.path.join(arcpy.env.scratchFolder,"outData.shp")‍ ‍ or outData = os.path.join(arcpy.env.scratchGDB,"outData")‍ and then arcpy.SetParameter(1,outData)‍ ... View more

Are you using an unfederated ArcGIS Server, or is it federated to Portal?  Your initial post indicates that you want your users to sign in with their Windows Domain credentials, indicating you're using AD.  If you're using an unfederated ArcGIS Server, you can configure two web adaptors with ArcGIS Server.  One will be used internally for users and can be configured with Windows Authentication to provide a single sign on experience.  The other can be used externally to still allow access to unsecure services.  Both can be external as well, but you'd want to be careful to create applications specific for the group of users accessing the application, (external users vs users with valid Windows credentials). ... View more

What type of authentication have you configured at the web tier level?  You should enable Windows Authentication so that you can use single sign on which eliminates the sign in screen altogether.  The credentials are passed in from the user accessing the application. ... View more