|
POST
|
Hello, I am trying to make a backup of our ArcGIS Enterprise configuration using webgisdr tool but it is failing to backup the hosting server (one portal, one federated server and 2 datastores work fine). I have the following error message: Failed to back up the ArcGIS Server: Url: https://gisags.enterprise.com/arcgis. Failed to generate a token to access ArcGIS Server from Portal. Setting webgisdr logging to debug, there is the following error: {"code":500,"messages":["Error while processing request. Internal error. JSON response not set."],"status":"error"} Setting the faulty ArcGIS Server log to DEBUG, I see the same error: "java.lang.Exception: Error while processing request. Internal error. JSON response not set" from "Admin" source. On portal log, I get the following errors: "Generate Server Token failed, serverId not found " Server URL can't be validated for granting server token 'https://gisags01.enterprise.com:6443/arcgis' Also, export arcgis server site manually from admin interface works fine: https://gisags.enterprise.com:6443/arcgis/admin/exportSite I don't know what is wrong with the configuration but when I validate the servers from portal administration interface both my hosting server and federated server are valid (green check mark). So what is webgisdr trying to do ? What is the request that fails ? Where should I look for ? Thanks, Nicolas EDIT: What I suspect: I configured an alias on the hosting ArcGIS Server ("gisags") while the actual hostname is "gisags01". I configured an SSL certificate on ArcGIS Server with both domain registered so both https://gisags.enterprise.com:6443/arcgis and https://gisags01.enterprise.com:6443/arcgis url are valid from an SSL point of view. On Portal for arcgis, it is the alias which is registered: https://gisags.enterprise.com:6443/arcgis and the validation from portaladmin works. But the portal log mentioned above makes me think that webgisdr found out the hostname and use it to generate its token which works but then it tries to find the id of the ArcGIS server from the portaladmin based on the server name and it does not find because its name is with the alias ("Server URL can't be validated for granting server token 'https://gisags01.enterprise.com:6443/arcgis'"). Description of the server from the portaladmin REST interface (https://myportal.enterprise.com:7443/arcgis/portaladmin/federation/servers/foobarid) Name: gisags.enterprise.com:6443 Id: foobarid Url: https://gisags.enterprise.com/arcgis Role: HOSTING_SERVER Admin Url: https://gisags.enterprise.com:6443/arcgis The hostname "gisags01" is not mentioned anywhere, so I believe it is why portal logged: "Generate Server Token failed, serverId not found. Server URL can't be validated for granting server token 'https://gisags01.enterprise.com:6443/arcgis'" I am only guessing... One may wonder, why did I use an alias: in the documentation for a highly available portal, it is written; "Note that when you federate a highly available ArcGIS Server site with Portal for ArcGIS, set Administration URL to a URL that the portal can use to communicate with all servers in the site, even when one of them is unavailable, such as a load balancer URL." (lien). I thought it would be more flexible to use an alias if later I want to add a load balancer or add another arcgis server to the site and kill the previous host for example. Is it a misconfiguration or rather a bug from the webgisdr ? Jonathan Quinn, I believe you might have the answer ?! Thanks for your help,
... View more
03-02-2019
10:49 PM
|
1
|
21
|
10445
|
|
POST
|
Hi, I just upgraded from 10.4.1 to 10.6 and now my script is failing with the same error: ERROR 000260: The table already exists.--The item was not found.--The item was not found. Failed to execute (Copy) No table has the same name and this script ran successfully on 10.4.1. Feature classes has attachments : BUG-000109584 BUG-000109584: The Copy tool does not execute successfully in ArcGIS Desktop when copying feature datasets with image attachments. Thanks
... View more
01-22-2019
11:39 PM
|
0
|
4
|
3686
|
|
POST
|
Yes better for me as well on desktop browers. Unfortunalety, my android device did make it through this release. Now the map is not loaded at all stating that: "WebGL support for high precision fragment shaders is required but not supported"
... View more
12-22-2018
04:35 AM
|
0
|
1
|
5602
|
|
POST
|
Thanks for the update. I was looking for the same functionnality. Too bad it is not supported specially with 4.X JS API which offer map rotation for free !
... View more
12-22-2018
04:32 AM
|
1
|
0
|
1755
|
|
POST
|
Actually I noticed the same thing as you on an Android device I have, using Google Chrome. It fails right away and features are not drawn. On desktop, it is working fine. WebGL joy
... View more
12-08-2018
02:22 AM
|
1
|
1
|
5602
|
|
POST
|
Thanks for the reply. Not in your priorities I guess. Good to know though.
... View more
11-29-2018
12:13 AM
|
0
|
0
|
6377
|
|
POST
|
Hello, I know directions widget of 4.9 is still in beta but I think it would be interesting to be able to remove the default ESRI world geocoder from the search sources of the widget just like what we can do currently with the property includeDefaultSources of the search. Currently, defining the sources of the search in the searchProperties property do not overwrite the default ESRI world geocoder but append it and I was not able to find an easy way to remove it as the searches are not "accessible" as a property of the widget. I tried to set my source as default with the activeSourceIndex property, but it seems like it is not taken into account. In the meantime, if you have any idea for a workaround, that would be great ! Thanks,
... View more
11-27-2018
11:21 AM
|
0
|
2
|
1666
|
|
POST
|
Hi, I do have noticed that as well with 4.9 but for me features are drawn anyway and it occurs out of the blue after playing a long time with the map: Uncaught TypeError: Failed to execute 'shaderSource' on 'WebGLRenderingContext': parameter 1 is not of type 'WebGLShader'. at t._loadShader (Program.js:25) at t.initialize (Program.js:25) at t.bindProgram (RenderingContext.js:25) at t.drawGeometry (WGLGeometryBrushLabel.js:25) at eval (WGLGeometryBrush.js:25) at Object.n [as forEachIter] (iterator.js:25) at t.draw (WGLGeometryBrush.js:25) at eval (WGLTile.js:25) at Array.forEach (<anonymous>) at e.doRender (WGLTile.js:25) I was not able to find a pattern... not always easy with WebGL...
... View more
11-27-2018
11:06 AM
|
0
|
3
|
5602
|
|
POST
|
Hello Bjorn Svensson, It seems like the issue remains in 4.9. Any hope for 4.10 ? Is there NIM for following the status ? It is a big a limitation for us as it prevents us from fully exploiting AGOL webmap philosophy. Thanks, Nicolas
... View more
11-19-2018
02:20 AM
|
0
|
5
|
6377
|
|
POST
|
Hello, We noticed a very high CPU utilization of our WebApp built with ESRI JS API 3.26 even when non interacting with it. Digging into it, it turns out that we included a custom widget which is an iframe application built with React Nativ making "window.parent.postMessage" call to talk to our WebApp in order to send some information. The problem is that the "esri/geometry/geometryenginewebworker.js" listens to this message as well and do a switch case according to the message sent (case "equals", "disjoint", "within", ...). Of course, as it is not concerned by this message, it goes to the default case which itself post another message "Unrecognised Action" and there we go, we run into infiinite loop : Can it be fixed please ? Any worker around that come to your mind ? Here is a code pen illustrating the problem: https://codepen.io/anon/pen/zMELJK If you remove the postMessage section from the codepen ("window.parent.postMessage("Hello",window.location.origin))", you should see your CPU calming down. Thanks,
... View more
11-16-2018
07:26 AM
|
0
|
0
|
967
|
|
POST
|
Hi Jack, Sorry, I had misunderstood your point. It does return some 404 errors for me as well. It is I believe the normal workflow: I request a tile which does not exist (404 error), resampling is enabled so OK let's resample the previous one ! What I meant on my previous post related to 404 errors, is that when reaching the level for which there were no tiles (in my case level 15) it tried to resample the previous one (14) but as my schema was not a level/2 one, it looked for a wrong tile so it returned 404 error once again and so on even for levels which had tiles. So at the end, I had a blank tile when "overzooming". If you pay attention to the 404 errors I had on my post, it goes from level 15 to 0 but tiles did not exist only at level 15 ! I hope it is clearer but I think you are good !
... View more
09-26-2018
03:55 AM
|
0
|
0
|
3193
|
|
POST
|
Not a bug ? Something wrong made on my side ? Cheers
... View more
09-26-2018
03:41 AM
|
0
|
3
|
1736
|
|
POST
|
Hi Jack, On my side 404 error was due to the tiling scheme not following a scale/2 pattern. Once I updated the tiling scheme, it worked. Also, I was using 3.25 JS API. Are you using 4.X by any chance ? Maybe the behavior is different. I never tested on 4.X. Would you have a url to test ? Nicolas
... View more
09-20-2018
06:13 AM
|
1
|
2
|
6265
|
|
POST
|
Thanks for the reply. Glad it has been logged as it was already there is 4.6 and 4.7.
... View more
07-23-2018
06:51 AM
|
0
|
0
|
3907
|
|
POST
|
Hello, I noticed that the search widget from JS API 4.8 fails if an integer field with exact search is configured: $ Here is a codepen based on js api sample: Search widget with multiple sources - 4.8 I just modified the search field to an integer one: "MEDHINC_CY". Search for 40421 and I will fail. Thanks
... View more
07-21-2018
07:07 AM
|
0
|
2
|
4363
|
| Title | Kudos | Posted |
|---|---|---|
| 1 | 2 weeks ago | |
| 2 | 3 weeks ago | |
| 4 | 06-04-2026 10:51 PM | |
| 1 | 05-30-2026 03:46 AM | |
| 2 | 05-30-2026 01:01 AM |