When using the backend to edit pages, eventually the Umbraco backend will cease to return correctly, which will eventually cause the whole website to crash.
This has happened ONLY when someone is working in the backend.
Each request not returned will lock up a CPU core (presumably stuck in an infinite cycle?) until it hits the timeout period of 100 seconds in which the process will forcefully close.
It is apparent here that it is internal Umbraco calls that is not returning out of those requests.
We are currently using 8 cores, and at this point in time you I can see that 4 cores are currently locked up.
I went on to open a new browser window and made it instantly to the Umbraco Login screen, entered my credentials, and hit Login. The request did not complete. I hit Login 4 more times, and as expected, all the cores locked up.
Immediately after all cores were locked up, requests came flooding in and were blocked due to all cores being occupied.
A quick restart of IIS fixed the issue back up, with the backend working correctly, and those same nodes were again functional.
Is there a way to avoid this from happening? There is currently no custom code in the backend that I have added, which leads me to believe it is an internal error, or that our database is somehow corrupt.
Umbraco Requests Crashing
Umbraco version 7.2.0
When using the backend to edit pages, eventually the Umbraco backend will cease to return correctly, which will eventually cause the whole website to crash.
This has happened ONLY when someone is working in the backend.
Each request not returned will lock up a CPU core (presumably stuck in an infinite cycle?) until it hits the timeout period of 100 seconds in which the process will forcefully close.
It is apparent here that it is internal Umbraco calls that is not returning out of those requests.
We are currently using 8 cores, and at this point in time you I can see that 4 cores are currently locked up.
I went on to open a new browser window and made it instantly to the Umbraco Login screen, entered my credentials, and hit Login. The request did not complete. I hit Login 4 more times, and as expected, all the cores locked up.
Immediately after all cores were locked up, requests came flooding in and were blocked due to all cores being occupied.
A quick restart of IIS fixed the issue back up, with the backend working correctly, and those same nodes were again functional.
Is there a way to avoid this from happening? There is currently no custom code in the backend that I have added, which leads me to believe it is an internal error, or that our database is somehow corrupt.
Would strongly advise an upgrade to 7.2.8. :)
https://our.umbraco.org/contribute/releases/728
is working on a reply...