I have an application pool configured to be have maximum 1 worker process.
Other settings on the pool:
- .NET : v4.0
- 32bit : True
- StartMode : AlwaysRunning
- Identity : domain\account
- Idle Time-out : 0
- Disable Overlapped Recycle : False
- Specific Time Recycle : 02h00
This morning I noticed that the CPU of the server is on 100% constantly.
2 weeks ago the same problem happened. I killed the process ID that was still on running, and a bit later there was only 1 pool left.
If I look to the worker processes, I see that there is one on state "running". It looks like the new worker was started after recycle. And it is now waiting for the old one to shutdown. But it's stuck somewhere since 24/02 ~ 25/02.
In Process Explorer on the process start timestamps. I see that the second and third worker started on the 25th. The weird is here that the 2nd is 3 seconds before the recycle config value. And the 3rd is exact on the recycle time.
And since then each day happens the same. At this moment there are 29 exe's instead of 1 for this pool.
In the EventViewer I have found no warnings/errors related to this application pool.
When I do a manual Recycle, the issue does not happen.
I have two web servers with a loadbalancer in front. And on both the servers the same issue happened.
I did some googling, but I have not found anything similar to this.
In the pools there are multiple WCF services running. This is installed on customers servers in their own network. But we have this at 50+ customers, and we only had this issue at this customer.
Windows Server 2016 Standard, 1607 (14393.3443) IIS Version 10.0.14393.0
