we are using wildfly-9.01 on azure environment and using application gateway and WAF services.
We observe 502 in apache log in underload and application became unresponsive.
We are using following configuration.
OS: centos-8(32 vcpus, 256 GiB memory)
web server: apache with modjk
application server: wildfly-9.01(40 GB memory is allocated)
Below is the Wildfly configuration:
<subsystem xmlns="urn:jboss:domain:io:1.1">
<worker name="default" task-max-threads="500"/>
<buffer-pool name="default"/>
</subsystem>
<subsystem xmlns="urn:jboss:domain:undertow:2.0">
<buffer-cache name="default"/>
<server name="default-server">
<ajp-listener name="ajpListener" scheme="http" socket-binding="ajp" max-header-size="50000" max-post-size="5368709120" max-parameters="5000000" allow-equals-in-cookie-value="true"/>
<http-listener name="default" socket-binding="http" redirect-socket="https" max-header-size="50000" max-post-size="5368709120" max-parameters="5000000" allow-equals-in-cookie-value="true"/>
<host name="default-host" alias="localhost"/>
</server>
<servlet-container name="default">
<jsp-config x-powered-by="false"/>
<websockets/>
</servlet-container>
<filters>
<response-header name="server-header" header-name="Server" header-value="My company"/>
<response-header name="x-powered-by-header" header-name="X-Powered-By" header-value="My company Application"/>
</filters>
</subsystem>
Below is the Apache Keep alive config:
KeepAlive On
MaxKeepAliveRequests 300
KeepAliveTimeout 600
TimeOut 1200
any suggestions for things to look at or try to avoid 502?
