Hi @musa.demir
No, nothing on server log file. Seems to be a client side issue only
Tks for your feedback. After lot of time lost on this issue, it was finaly a very simple one ! Just forgot to close <head> tag on index html file... since 2 years now :D Update from Angular 10 to 11 make this bugg visible
Hi @ismcagdas,
Tks for your reply, do you have further information to share ?
Hi @ismcagdas,
Did you had time to investigate ?
@ismcagdas, I've sent an email with all details for your investigation.
Hi @ismcagdas,
Yes I'm in a single solution and hosting an angular app. Therefore, seems that I need that line :)
But I'm having a lot of logs about it and it seems to be linked with SignalR disconnections. I already post a question about similar issues in the past... Right now, I'm ignoring these errors.
The problem is that I noticed a huge CPU consumption on my Azure server due to 404 failed requests on SignalR :
And the only error logs that I have is the ones about this code line (line 166 in startup.cs).
I also found this ABP issue that might be the same error : https://github.com/abpframework/abp/issues/1830
Do you think that it could be linked to the SignalR reconnection startegy on clients (angular app or Xamarin app) ?
Hope you can help me on that because the risk is high when a lot of users are using the app at the same time.
Hi @ismcagdas,
I will try to convert to a thread less method and add a lock. Thanks
Hi maliming, Not so much, arround 6000 rows
Hi @ervingayle,
A user pointed out an issue while saving an entity with specific data. After analysis in debug, we figured out the app crash leading to 100% CPU consumption. I think the problem is wihtin your code. Try to look at your specific code in comparison with zero code basis ; and be careful with EF Core 3.x...
Hi,
I had an Azure support ticket opened untill yesterday. We monitored the app actively and found several issues.
The most important one was not logged : it was a crash due to EF Core 3.1. My app has ben migrated to ASP.NET Core 3.1 some weeks ago ; as you know, some refactoring is needed due to EF Core breaking changes... Despite all of our tests during this migration, we miss to test one special case where the app was deleting database items > we noticed the app crashed wihtout any logs causing a 100% CPU due to app restart ! By replacing EF method DeleteAsync by HardDeleteAsync, we fixed this BIG issue. The only clue that we had from the monitoring was an exception 0xC00000FD Stack overflow... Of course, before finding this issue, we search within the app logs to find something strange ; and as you also mentioned, a lot of ERROR and WARNING logs are not affecting the app at all. We lost a lot of time arround them :(
This was the root cause of this ticket. Honestly, we lost a lot of confidence on our app as this was happening in production and affecting all users. Particularly on EF Core side and Log management.
We also fixed some other issues arround SignalR specially on client side for reconnection (following ASP.NET Zero code base).
I'm closing this ticket as it is resolved from our side.
@ervingayle, if you have any question, feel free to ask :)
