We’ve been testing the expense management app in DEV. It has its own AppServer (on prem) set up for basic auth only and it’s accessed through a non-standard port.
We lately added a 3rd company to the same dev environment and that seems to work, but coincidentally or no the expense app now crashes a second after logging in.
The server logs: Microsoft.AspNetCore.Connections.ConnectionResetException: The client has disconnected ---> System.Runtime.InteropServices.COMException (0x80070040): The specified network name is no longer available. (0x80070040)
At first it returned a token error to the user until I deleted and re-registered the appserver.
Then it hung on login and the server logs pointed out that the service accounts weren’t valid for the new company.
Now I’ve made sure the new company is added to the users, the service accounts, and the manager account that runs things. I’ve regenerated the data model and made sure all the separate appservers have had their app pools recycled. Finally, I rebooted the server.
Try to catch what happens in browser, when it crashes.
Open Developer tools, check the checkboxes that save console and network logs between calls.
Reload page in browser and see if any outstanding error shown. Maybe if you have working instance, compare log with it.
Thanks @Olga ! It doesn’t crash in the browser, only in the app, but If I understood you correctly, I logged into Kinetic in the browser, opened dev tools, and ticked “preserve log” on the Network tab, and then tried to log in on the app. I see some errors but I think they’re all from the browser login.
I am not familar with that app, but that error that you show
The client has disconnected —> System.Runtime.InteropServices.COMException (0x80070040):
can be valid error when server call was canceled from browser.
So it looks like something happens during login - it cannot process response and resets to default page. Test mode is just a default badge value for Kinetic UI library.
So probably errors from browser login may help to understand what is not right.
Also try to use incogito mode in browser and check if it changes something.
I was just getting the same issue on an install. I had forgotten to add my Server Service account that is used for the app pool, to the SQL Server Management studio as a user sysadm server role.
Tried to do a trace, both in the client and in admin console. Nothing but User Tracing tags.
Anyone know how to do a trace on mobile app logins?
Edit - ok, found it here:
But still stumped
Edit: okay, interestingly I CAN see the restapi calls in a trace, but not sure if I’m seeing the responses. Nothing looks like JSON or response codes but I’m way outside my comfort zone.
Edit: I can enter and submit an expense in Kinetic, in the browser, on my phone. So the actual connection is working; the mobile app, however, is not. Support ticket open for almost a week and all I’ve been able to do so far is send availibity and decline a meeting invite that wasn’t in that availibility
Response codes can be found in IIS log, but basically if there is no exception in the log, they are 200, with exception - 500. JSON is not dumped in the log.
Hello all, interesting closure on this one. We suspected all sorts of things (DNS, Bindings, SSL, etc.) and Epicor support ended up being very helpful. Ultimately, however, today when I opened the app, while on a call with support to try to figure this all out, the app looked subtly different from a couple of days ago.
It also worked.
I take from this that 99% of the app doesn’t live on my phone, because my phone doesn’t accept updates without me agreeing.
The support rep was able to confirm when he looked into it that there was an update a few days ago. I assume it took a few days for CDNs to replace their cache, maybe?
Sadly, in the time it took, our finance department more or less gave it up as a bad job, but in view of an integrated UX I’m going to try to re-evangelize this for them.
No idée, I’m grasping at straws!