I am running an apache httpd server, v2.2. It has been working for some time, but since the last restart (the VM had to be restarted), connections (for example from the same machine with wget) to it simply time out.

netstat -tulpn has shown the process is listening on port 443 as expected. The logs do not show anything (although I have set the LogLevel to debug). Also I noticed that many zombie processes are spawned and killed again -- always 1 to 5 "defunct" processes are shown when I do ps -elf | grep http.

Any ideas about what may cause this? The configuration has not changed.

  • 1
    Ate you using any other module like php or fast_cgi ?
    – user34720
    Jan 18, 2017 at 9:40
  • mod_shib and mod_jk Jan 18, 2017 at 9:52
  • Perhaps firewalld or iptables block port 443? (Although it would be surprising that they also filter connections from localhost.)
    – dr_
    Jan 18, 2017 at 9:56
  • what is the load of the server? Are there any unexpected logs in syslog? Logs of tomcat/catalina.out? mod_jk is normally answering the requests. Jan 18, 2017 at 10:34
  • Nearly no load. Nothing at all in the logs. Thanks for the mod_jk hint, I will configure the log level of mod_jk to debug and try again. Jan 18, 2017 at 11:38

2 Answers 2


You wrote:

The logs do not show anything (although I have set the LogLevel to debug)

Make sure you do not override it later somewhere else in the configuration.

If it does not show any logging try to find out why it cannot write logs at all using something along the lines of:

sudo strace -f /path/to/httpd_start_script 2>&1 | less -i
  • Thanks! The actual issue was very specific and has been solved (see my answer), but I'll keep this in mind. Jan 18, 2017 at 15:20

It turned out that wrong permissions have been set to the shibboleth configuration file, which caused the mod_shib module to not work, in such a way that no logs were produced.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .