Problem:
During the performance test, the server stopped responding after 10 users continued to run for several minutes concurrently.
From the LR error point of view, there are the following:
1) error-27728:step download timeout (seconds) has expired when downloading Non-resource (s)
2) Error-27791:server "Appsvr01" has shut down the connection prematurely [msgid:merr-27791]
3 error-27796:failed to connect to server "appsvr01:38080": [10061] Connection refused
4) Error-26372:contentcheck rule "StackTrace" in Application "Bioffice" triggered. Text "", "StackTrace": "" Matched (count=1) [msgid:merr-26372]
But the latter two should be indirect errors caused by the previous two errors. In the order of errors, the previous timeout error occurs first, and the entire server stops responding.
Positioning:
1 connect the server with the client, no response;
2) netstat see the server port is still listening, but telnet the port is forced to disconnect, indicating that webserver has not been normal;
3 from the server log to see a outofmemory error, the judgment is related to memory leaks.
Reproduction and Analysis:
1 set-verbose:gc-xloggc:./gc.log print GC log, rerun the script, reproduce the problem.
When you start running from the GC log, garbage collection occurs at regular intervals, but the memory stack is used more and more, the last memory is not enough, and the server stops responding with frequent full GC.