Pi-hole Web Interface unavailable, Error 503 Topic is solved

[pwrdbytrd]

Since updating to the most recent version if Dietpi I can no longer access my web interface. I completed removed and reinstalled pi-hole via the dietpi-software option. I can control pi-hole via ssh but I cannot access the webpage. I receive a xheader error when running pihole -d to view the debug log. I also am unable to upload the debug log as it asks at the end of the debug log. I am a novice with regards to Linux and Dietpi but have really enjoyed this feature. I didn't make any other changes other than updating the dietpi software so I'm not really sure what could've happened. Any help is appreciated!

[pwrdbytrd]

Any help would be appreciated, I'm still at a standstill after reaching out to Pi-hole.

[johnvick]

Can you check the relevant files and folders under /var/www have the correct ownership i.e www-data?

In my case they are /var/www/admin and /var/www/pihole

Is the webserver otherwise working?

[pwrdbytrd]

/var/www/pihole and /var/www/admin are both showing owner and group as www-data.

The only other web interface I use is for Transmission and it seems to be working correctly as I am able to log in and add a torrent.

[johnvick]

Try
# pihole -up

If that doesn't fix it it could be a php problem - try reinstalling php through the dietpi-software menu.

[pwrdbytrd]

Pihole -up, all up to date.

PHP -- uninstall, reboot, reinstall, reboot. Same result of error 503. Pihole developer also pointed to php being the issue. Any other ideas?

2019-06-01 16:26:17: (mod_fastcgi.c.2686) fcgi-server re-enabled: unix:/run/php/php7.3-fpm.sock
2019-06-01 16:26:54: (mod_fastcgi.c.1926) connect failed: No such file or directory on unix:/run/php/php7.3-fpm.sock
2019-06-01 16:26:54: (mod_fastcgi.c.2842) backend died; we'll disable it for 1 seconds and send the request to another backend instead: reconnects: 0 load: 1
2019-06-01 16:26:54: (mod_fastcgi.c.1689) all handlers for /index.php? on .php are down.

[pwrdbytrd]

Please see attachments below.

I have now gotten to a point where if I enter the command

Code: Select all

systemctl start php7.3-fpm

the web interface will work correctly until I reboot.

If I reboot and run

Code: Select all

service php7.3-fpm status

it shows

Code: Select all

Active:  failed (Result: timeout)

[johnvick]

Any clues in the php logs?

[MichaIng]

@pwrdbytrd
PHP-FPM times out on boot. Looks like it is related to the low entropy issue: https://github.com/MichaIng/DietPi/issues/2806
Real source and solution for this issue is outside of our control, seems to need to fix on Debian side. But you can work around it by installing haveged: G_AGI haveged

While this solves hanging boot and service start timeout, it is not a very well solution since it bears the risk of pseudo-randomness which is a potential security risk when you think about encryption keys where it is essential that /dev/random provides real random output that is not predictable or something.

Since I want to collect info about which systems are affected, might you please send us a bug report and paste the ID here?
dietpi-bugreport

[pwrdbytrd]

@pwrdbytrd
PHP-FPM times out on boot. Looks like it is related to the low entropy issue: https://github.com/MichaIng/DietPi/issues/2806
Real source and solution for this issue is outside of our control, seems to need to fix on Debian side. But you can work around it by installing haveged: G_AGI haveged

While this solves hanging boot and service start timeout, it is not a very well solution since it bears the risk of pseudo-randomness which is a potential security risk when you think about encryption keys where it is essential that /dev/random provides real random output that is not predictable or something.

Since I want to collect info about which systems are affected, might you please send us a bug report and paste the ID here?
dietpi-bugreport

00b60c5e-4a10-48a5-8a2b-a9c5c950a28c