Its noted that i am getting a new error all of a sudden in the Web UI “Backend icinga not running”.
I have done configuration check using icinga2 daemon -C and could not find any relevant issue related to this.
Can someone help me with this.
Along with this, i am attaching the icinga2 and message logs for references.
i have ran “select status_update_time from icinga_programstatus;” to check if DB is getting updated. Its also showing the latest time. Attaching the snap for your references.
in the log files you provided I found the following entry:
SELinux is preventing /opt/rh/rh-php71/root/usr/sbin/php-fpm from name_connect access on the tcp_socket port 3306.
3306 is the default MySQL/MariaDB port. I think SELinux is blocking the php-fpm daemon to access the IDO database. I am not familiar with SELinux, but you can try to disable it for testing purposes and see if the error message “backend is not running” will disappear. If SELinux is the problem you can investigate how to create a exception for this.
for checking the logs about SELinux issues, you better should look into /var/log/audit/audit.log. Check here if you find some entries with “denied”.
Sometimes you could find also entries in the journal. Here you could grep for “setrouble”.
Our experience is very often that if you fix one issue, the next one is comming. So we had to fix one permission issue at a time until it worked. We we knew all needed permissions, we create or own SELinux rule set if there is nothing delivered with the installation routine.
Can you please test if disabling SELinux will solve the issue. Run setenforce 0 to disable the enforcing mode, now restart Apache and/or PHP-FPM and look if the issue is resolved. If so we know that this is caused by SELinux and can look deeper into it how to resolve this with enabled SELinux. If the issue is not resoled we know there must be another issue somewhere.
Please check the time/NTP settings on all hosts. Icinga Web is checking timestamps and if you hosts (including your client) think they are in different realities with different times, it can sum up to Icinga Web 2 thinking it doesn’t updated.
Also, check if you actually get new results. So is it about Icinga Web really not being updated or is it just this warning?
Please don’t just check the timezone but also if all systems show the same time. If e.g. ntp or chrony are not running, the could have different time settings. Please log into all systems and use date and ntpq -p to show if they show the same time and if NTP is synchronized.
zones.conf (1.1 KB) zones.conf (1.1 KB)
