Fresh Rocky 9.5 ISPConfig 3.2 install - ISPConfig interface won't load

using the "The Perfect Server CentOS 8 with Apache, PHP, Postfix, Dovecot, Pure-FTPD, BIND and ISPConfig 3.2" install guide as the base instructions, slightly modified for v9.5 (as per some of my posts on that guide already indicate). Zero issues with the installation on v9.5 other than the amavisd-new needing to be uninstalled/reinstalled/reenabled after ISPConfig was installed
EXCEPT...
after the ISPConfig was installed, the interface for it will not load. httpd was working when I installed RoundCube right before ISPConfig, so something in the ISPConfig install appears to have broken something. I did a --force reinstall of ispconfig3 after the initial install because I forgot to set up a DNS entry for this new server (so the LetsEncrypt failed), and I accepted all the defaults when doing the reinstall other than recreating the SSL certificate.
here's the TAIL of the httpd/access_log showing roundcube connections, then me trying to access ISPConfig @ http://192.168.0.123:8080

Code:

192.168.0.22 - - [19/Mar/2025:01:39:29 -0400] "POST /roundcubemail/installer/index.php?_step=3 HTTP/1.1" 200 4868 "http://192.168.0.123/roundcubemail/installer/index.php?_step=3" "Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:136.0) Gecko/20100101 Firefox/136.0"
::1 - - [19/Mar/2025:01:59:49 -0400] "GET / HTTP/1.1" 403 7620 "-" "Mozilla/5.0 (ISPConfig monitor)"
::1 - - [19/Mar/2025:02:00:01 -0400] "GET / HTTP/1.1" 403 7620 "-" "Mozilla/5.0 (ISPConfig monitor)"
::1 - - [19/Mar/2025:02:05:02 -0400] "GET / HTTP/1.1" 403 7620 "-" "Mozilla/5.0 (ISPConfig monitor)"

and here's the TAIL of the error_log

Code:

[Wed Mar 19 02:11:42.087391 2025] [lbmethod_heartbeat:notice] [pid 1368:tid 1368] AH02282: No slotmem from mod_heartmonitor
[Wed Mar 19 02:11:42.094757 2025] [:notice] [pid 1368:tid 1368] mod_python: Creating 8 session mutexes based on 0 max processes and 25 max threads.
[Wed Mar 19 02:11:42.094773 2025] [:notice] [pid 1368:tid 1368] mod_python: using mutex_directory /tmp
[Wed Mar 19 02:11:42.115316 2025] [mpm_event:notice] [pid 1368:tid 1368] AH00489: Apache/2.4.62 (Rocky Linux) OpenSSL/3.2.2 mod_fcgid/2.3.9 mod_python/3.5.0 Python/3.9.21 configured -- resuming normal operations
[Wed Mar 19 02:11:42.115338 2025] [core:notice] [pid 1368:tid 1368] AH00094: Command line: '/usr/sbin/httpd -D FOREGROUND'
[Wed Mar 19 02:15:02.077258 2025] [autoindex:error] [pid 1457:tid 1614] [client 127.0.0.1:52186] AH01276: Cannot serve directory /var/www/html/: No matching DirectoryIndex (index.html,index.php,index.html,index.cgi,index.pl,index.php,index.xhtml,index.htm,standard_index.html) found, and server-generated directory index forbidden by Options directive
[Wed Mar 19 02:20:01.744603 2025] [autoindex:error] [pid 1457:tid 1615] [client 127.0.0.1:37598] AH01276: Cannot serve directory /var/www/html/: No matching DirectoryIndex (index.html,index.php,index.html,index.cgi,index.pl,index.php,index.xhtml,index.htm,standard_index.html) found, and server-generated directory index forbidden by Options directive

and here's the tail end of the httpd.conf file

Code:

# Supplemental configuration
#
# Load config files in the "/etc/httpd/conf.d" directory, if any.
IncludeOptional conf.d/*.conf
RequestHeader unset Proxy early

NameVirtualHost *:80
NameVirtualHost *:443
Include /etc/httpd/conf/sites-enabled/

Thoughts on what might be the problem?

 

connection timed out - nothing loads

 

I was actually setting up two servers simultaneously (same commands on each) for a Master and Slave. hadn't installed ISPConfig on the slave yet, so I compared the httpd.conf files and the Master had just the following three lines added to the httpd.conf file:

Code:

NameVirtualHost *:80
NameVirtualHost *:443
Include /etc/httpd/conf/sites-enabled/

and i just tried accessing the roundcube install on the slave server, and it's not responding now either.
i'm thinking the roundcube config that gets generated as the last step of that install might be corrupt. there is no closing ?> in the content they have you add to the file.

 

well adding that didn't fix the issue.

 

PHP tags don't need to be closed if the document only contains php.

https://www.php.net/manual/en/language.basic-syntax.phptags.php

 

How do you access the CP? Via http or https? And have you issued a SSL-certificate with the ISPConfig installer?

Also make sure that you have set selinux to permissive mode and not enforcing and as Till already suggested make sure that the firewall (or firewalld) doesn't block traffic on port 80/443/8080.

 

selinux is disabled
what I did figure out is that flushing iptables got it working, but it's not clear why
iptables -L before the flush:

Code:

Chain INPUT (policy ACCEPT)
target     prot opt source               destination
ACCEPT     all  --  anywhere             anywhere             state RELATED,ESTABLISHED
ACCEPT     icmp --  anywhere             anywhere
ACCEPT     all  --  anywhere             anywhere
ACCEPT     tcp  --  anywhere             anywhere             state NEW tcp dpt:ssh
REJECT     all  --  anywhere             anywhere             reject-with icmp-host-prohibited

Chain FORWARD (policy ACCEPT)
target     prot opt source               destination
REJECT     all  --  anywhere             anywhere             reject-with icmp-host-prohibited

Chain OUTPUT (policy ACCEPT)
target     prot opt source               destination

iptables -L after the flush

Code:

Chain INPUT (policy ACCEPT)
target     prot opt source               destination

Chain FORWARD (policy ACCEPT)
target     prot opt source               destination

Chain OUTPUT (policy ACCEPT)
target     prot opt source               destination

is that I only issued the following command to make sure I didn't get myself blocked on the local network

Code:

iptables -I INPUT -s 192.168.0.0/24 -j ACCEPT

so i'm not sure how things weren't working
/var/www/html is empty by default - i added an index.htm file there on one server to test it - that's appearing. and the roundcube login appears when request.
on the other server, the apache default page is appearing now.
so I guess I'll add an IPTABLES --flush to my instructions just to make sure something didn't get mucked up at the end of the install.
Thanks for the early morning (here anyway) help - much appreciated.

 

Is firewalld running? That's the default firewall manager which is enabled since EL8 by default.
You must either disable it, or use firewalld to configure your firewall.

For more info see: https://docs.redhat.com/en/document...tionality#deprecated-functionality_networking

 

closing this, but i'll answer the last question. i disable firewalld right after the minimal install is done - i like iptables personally.
that machine has been wiped and reinstalled again - working without issue this time. I'm guessing there was a typo somewhere that left something amuck. install is clean now though.
again, thanks for the help. in this case, it was easier to start over as there was nothing configured on it yet that was important, so better to make sure it was clean to start with.