My thingworx server(deployed in a virutal machine) beomce very slow suddenly today (it has running for months smoothly) . Whatever action you made in the thingworx composer, will cause the cpu usage up to 100% and it's occupied by the tomcat process, the memory usage is 65%, that's fine.
Meanwhile, in the Monitor->connected users , I notice there're more than 100 Administrator users. I didn't notice how many Administrators list there , but I'm sure only one Administrator connected when the cpu is 100%. (please check the attached picture). I've tried to terminate the other connected Administrator, but they appear again constantly.
I've tried to restart the tomcat and the VM , but it doesn't help.
Any idea about it ?
Thingworx 8.4.4 + Postgresql 10.x + Windows server 2016
The problem has gone.
I just found that the login user is always "Administrator" even if I login in with another user actually. So I cleaned the Chrome's cache, then the CPU usage of tomcat and the system response become normal. (Chrome is in my host machine while tomcat deployed in VM machine ). The connected users still list lots of Administrator, but it doesn't affect the performance.
Although the problem has gone , I believe there's a potential bug. I updated the jstack log which was saved before solving the problem, the log was recorded when opening a thing's service and the CPU was 100% at that time. Hope the development team can analyze it to find the root cause.
Please test this with another browser to see if the issue still occurs. We have had other reports of this under Chrome.
Also, are you using any Chrome extensions that may be affecting the issue?
Just wondering if you have had time to test with a different browser to see if you still see multiple instances of the Administrator user under Connected Users. Since we've seen this problem with Chrome in the past, it would be helpful to know if it extends to other browsers.