I opened up my webinterface and saw this.
The domains have been blurried out by me don't worry.
I did
pihole status
which gave me this:
root@raspberrypi:/home/pi# pihole status
::: DNS service is running
::: Pi-hole blocking is Enabled
I've noticed that the memory usage was at 79% so i did
top
which showed that only about 30% of the memory was being used.
I then did
pihole -d
my token is: ymte5nkich
I then rebooted my pi which fixed everything.
I had a look at your debug log (the part of /var/log/lighttpd/error.log that has been uploaded) and found a lot of lines like
PHP Warning: exec(): Unable to fork [grep -c ": query\[A" /var/log/pihole.log] in /var/www/html/admin/scripts/pi-hole/php/data.php on line 484
which means that the resource available to PHP have been exhausted. This also explains both of the things you were seeing:
htop rather than top to display the actual memory blocked by running applications).Currently, every element on the web interface is computed differently and apparently the statics part was not able to run properly, reporting zeros. This will be solved by FTL which will serve all data displayed on the web interface.
Are running other things beside Pi-hole on your Raspberry? I don't see any indication that the high memory usage could have been due to Pi-hole alone in the debug log.
I don't run anything else on my pi.
I'm wondering how the resources that are available for PHP can be exhausted since you just raised the available memory to 256MB (if i remeber correctly).
Hmm, you might want to look around for the PHP error shown above. I can only guess what might have happened here.
Looking through my error.log didn't helped me much. 95% of the log is "Unable to fork".
It does not help me much if the log just tells me that it wasn't able to get the data.
Well this is not exactly what is happening here - it doesn't tell you that it was unable to get the data, it tells you that PHP wan't able to fork on exec, i.e. it was not able to run the specified commands. If that happens again, you could post the output of
cat /proc/meminfo
and
iotop
(the latter one might not be installed by default)
I'll remember it.
As soon as this happens again i'll be here.
I have pi-hole installed on a raspberry pi 3 along with Volumio.
The add blocking works (as far as I can tell) but the interface is all zero, like there were no requests to pi-hole.
Only things different are the following: in /etc/lighttpd/lighttpd.conf I have
server.port = 8080 (since on port 80 there is Volumio interface running)
server.errorlog = "/var/log2/error.log" (because lighttpd complained about not being able to access the log files)
accesslog.filename = "/var/log2/access.log"
Immediatly after the install the interface showed querries, but after restart only shows zeroes.
I generated a debug log and the token is e5dd0m6pza
If anyone can point me in the right direction about getting the statistics back, I thank you in advance!
did you disable logging when you installed also what OS are you running some flush log files in an attempt to save space
I haven't (voluntarily, at least) disabled logging.
System is Raspbian GNU/Linux 8 (jessie)
I realized the problem: lighttpd does not have access to the logging dir and to the pihole.log
So the solution I found is the following:
sudo mkdir /var/log/lighttpd/
sudo setfacl -m u:www-data:rwx /var/log/pihole.log
sudo setfacl -m u:www-data:rwx /var/log/lighttpd/
sudo /etc/init.d/lighttpd start
problem is after the pi reboots needs to be applied again.
I tried to put it in /etc/rc.local and /etc/init.d/rc.local
but it doesn't work in any of those two places although I read this is the place to put a commands to be run at startup as root...
Now I solved it!
last problem was that I read that rc.local content is executed as root and I omitted the sudo in front of the commands.
Now I put it back and it works!
running pihole -r and selecting the repair option should restore any missing permissions permanently as well
