Pihole slows to a crawl unless I flush constantly

The web interface for my pi-hole instance stops responding after ~6 hours of uptime. Running pihole -f will repair the web interface, but even SSH'ing into the container response times have slowed to a crawl for any commands.

After running pihole -f, the web interface will come back up and everything works fine. Running pihole status while it is in this state shows no failures and dns seems to work fine.

Is pihole supposed to flush automatically? How big do the log files tend to get? I will try and do a df -h on the box next time, perhaps it is filling the disk.

This debug was taken after a fresh flush, I can try to provide one while it is barely running.

Debug Token:


Does the same thing happen when you disable Conditional Forwarding? Forwarding often creates a loop that will crash things if it's not done correctly.

Edit: Yes, I see you have Pi-hole forwarding to for and then there are queries coming in from to Pi-hole.

Thank you for the quick response!

I have turned off conditional forwarding... I now see how that could be an issue if the requests aren't specifically tagged. Is there another easy way to name my clients? I'm forced to do a lot of troubleshooting right now since I just setup lancache and had to figure out how to integrate that with pihole (would make a great addition to pihole to easily integrate that!). As well, trying out Peacock and it freaks out if ads are blocked in stream, so trying to track down requests from clients would be a lot easier with named clients.

Consider using Pi-hole as the DHCP server.

I was very tempted to on setup but was concerned about uptime/stability since this is running on my main server which can go down from time to time either for planned upgrades or unplanned failures and wanted the rock solid experience my router provided. I may need to reconsider it.

This topic was automatically closed 21 days after the last reply. New replies are no longer allowed.