search

chriscrowe / docker-pihole-unbound

Run Pi-Hole + Unbound on Docker
1.06k stars 275 forks source link

Rate-limiting 10.0.2.100 #124

Closed kostadinnm closed 2 years ago

kostadinnm commented 2 years ago

Hello. I've been utilizing the single-container setup at home. It's all in a podman container on an an almalinux(centos-based) homelab box. On my personal laptop I've just tried to restore a couple of docker containers, which auto-pulls some images and I started getting timout errors stating 60 seconds are exceeded. I understand that pi-hole had enabled rate limiting for a while now. The question is shouldn't the pihole instance be excluded from this limitation and not change the limit for all the nodes in my network?! As I am a power user, I might bump the limit, but how should I know how many requests is my use case making?! Here's some of the ftl log and it doesn't state the exact query exceed number:

[2022-04-21 18:20:59.069 521M] Rate-limiting 10.0.2.100 for at least 36 seconds
[2022-04-21 18:21:35.045 521/T538] Ending rate-limitation of 10.0.2.100
[2022-04-21 18:27:33.990 521M] Resizing "FTL-queries" from 688128 to (16384 * 56) == 917504 (/dev/shm: 1.2MB used, 65.5MB total, FTL uses 1.2MB)
[2022-04-21 18:27:34.514 521M] Rate-limiting 10.0.2.100 for at least 1 second
[2022-04-21 18:27:35.496 521/T538] Ending rate-limitation of 10.0.2.100
[2022-04-21 18:28:02.208 521M] Rate-limiting 10.0.2.100 for at least 33 seconds
[2022-04-21 18:28:35.571 521/T538] Ending rate-limitation of 10.0.2.100
[2022-04-21 18:31:28.756 521M] Rate-limiting 10.0.2.100 for at least 7 seconds
[2022-04-21 18:31:35.800 521/T538] Ending rate-limitation of 10.0.2.100
[2022-04-21 18:35:47.594 521M] Rate-limiting 10.0.2.100 for at least 48 seconds
[2022-04-21 18:36:35.176 521/T538] Ending rate-limitation of 10.0.2.100
[2022-04-21 18:43:56.125 521M] Resizing "FTL-dns-cache" from 8192 to (768 * 16) == 12288 (/dev/shm: 1.4MB used, 65.5MB total, FTL uses 1.4MB)
[2022-04-21 18:45:18.744 521M] Resizing "FTL-queries" from 917504 to (20480 * 56) == 1146880 (/dev/shm: 1.4MB used, 65.5MB total, FTL uses 1.4MB)
[2022-04-21 18:45:22.914 521M] Rate-limiting 10.0.2.100 for at least 13 seconds
[2022-04-21 18:45:35.862 521/T538] Ending rate-limitation of 10.0.2.100
[2022-04-21 18:53:49.110 521M] Rate-limiting 10.0.2.100 for at least 46 seconds
[2022-04-21 18:54:35.533 521/T538] Ending rate-limitation of 10.0.2.100
[2022-04-21 18:58:52.371 521M] Rate-limiting 10.0.2.100 for at least 43 seconds
[2022-04-21 18:59:35.908 521/T538] Ending rate-limitation of 10.0.2.100
[2022-04-21 19:00:26.784 521M] Rate-limiting 10.0.2.100 for at least 9 seconds
[2022-04-21 19:00:35.980 521/T538] Ending rate-limitation of 10.0.2.100
[2022-04-21 19:02:33.241 521M] Resizing "FTL-queries" from 1146880 to (24576 * 56) == 1376256 (/dev/shm: 1.7MB used, 65.5MB total, FTL uses 1.7MB)
[2022-04-21 19:12:33.122 521M] Resizing "FTL-dns-cache" from 12288 to (1024 * 16) == 16384 (/dev/shm: 1.9MB used, 65.5MB total, FTL uses 1.9MB)

I've connected to a hotspot and downloaded most of the stuff(uncached dns-records, I guess) and it's ok now, but still this might be helpful to someone.

Cheers, Kostadin

kostadinnm commented 2 years ago

I think there might have been an (cross-) issue with my workstation machine and pi-hole itself. My work containers were re-launched with the latest images, but I continued to experience laggy execution of ping requests to arbitrary domains. This had been noticable in general internet browsing as well. Directly pining resolved ips was fast and flawless. I am closing this, as it may actually not be relavant here and I got it working for me without any modifications since. If it happens again, I'll try gather more context and probably report back(here or over to pihole/unbound ground).