[
https://issues.jboss.org/browse/WFLY-7457?page=com.atlassian.jira.plugin....
]
Stuart Douglas commented on WFLY-7457:
--------------------------------------
Is it possible that the F5 was not actually dropping the TCP connection properly? If no
timeout was configured then Wildfly may run out of memory as it will keep the data
structures associated with those connections around forever.
UT005013 causes OutOfMemoryError
--------------------------------
Key: WFLY-7457
URL:
https://issues.jboss.org/browse/WFLY-7457
Project: WildFly
Issue Type: Bug
Components: Web (Undertow)
Affects Versions: 10.1.0.Final
Reporter: Steve Vater
Assignee: Stuart Douglas
Labels: new_and_noteworthy
There's some F5 WAF within the network where our WildFly is running. The WAF sends
multiple Keep Alive-Requests to our WAF via TCP, but without sending
"close_notify" after the communication has ended.
Undertow throws UT005013, which is correct, but this causes the JVM to run out of Memory.
The keepalive is sent 12 times per minute (every 5 seconds) to our WildFly.
We already installed an update of Undertow (1.4.3); This didn't help.
--
This message was sent by Atlassian JIRA
(v7.2.3#72005)