new install v0.20.3 -> Could not resolve host.

[KGBist2000]

Describe the bug
New Clean install v0.20.3. A speedtest is failed everytime by following error. "Could not resolve host."
Container is started well without any issues. From container bash i can reach the internet by ping google.com without any issues.

To Reproduce
Steps to reproduce the behavior:

1. Go to 'Admin Page'
2. Click on 'Run Speedtest'
3. See error -> Could not resolve host.

Expected behavior
Start speedtest

Environment (please complete the following information):

• OS: Proxmox LXC, Portainer Docker
• Architecture: [ amd64]
• Browser [firefox]
• Version [v0.20.3]

Docker Compose

version: '3.4'
services:
speedtest-tracker:
container_name: speedtest-tracker-new
ports:
- 5443:80
#- 8443:443
environment:
- PUID=1000
- PGID=1000
- APP_KEY=base64:----- replaced ------
- DB_CONNECTION=sqlite
- SPEEDTEST_SCHEDULE="9 * * * *"
- SPEEDTEST_SERVERS="53128,48042,31469"
- APP_TIMEZONE=Europe/Berlin
- DISPLAY_TIMEZONE=Europe/Berlin
- TZ=Europe/Berlin
- APP_NAME="Speedtest"
- PUBLIC_DASHBOARD=false
volumes:
- /root/speedtest_tracker:/config
#- /path/to-custom-ssl-keys:/config/keys
image: lscr.io/linuxserver/speedtest-tracker:latest
restart: unless-stopped

[bigrichwood]

Same with my container which was upgraded a few hours ago from 0.19.0 to 0.20.3. Connectivity looks fine from within the container, but all tests fail with “could not resolve host”.

I’ve rolled back to 0.19.0 which works perfectly, and I’ll just leave it pinned to that version as it does everything I need.

[c11umw]

I wondered why it stopped working today.... that'll be it (watchtower auto-update)
Found this topic, forced 0.20.2 and speedtest started working again.

From within the 0.20.3 container I could successfully nslookup domains, run the php command to see the closest servers and curl https://ipinfo.io/ip which returned my WAN IP to prove internet and resolving was all working.

[alexjustesen]

v0.20.3 changes the internet check from a HTTP get request to a true ping. In some ISP cases it seems those requests could be blocked.

You can also change the default ping IP/URL which is Cloudflare's 1.1.1.1 by setting SPEEDTEST_PING_URL as an environment variable.

[c11umw]

Tested.

docker run -d \
--name=speedtest-tracker \
-e PUID=1000 \
-e PGID=1000 \
-e SPEEDTEST_SCHEDULE="26 * * * *" \
-e SPEEDTEST_PING_URL=1.1.1.1 \
-e PRUNE_RESULTS_OLDER_THAN=30 \
-e DB_CONNECTION=sqlite \
-e APP_KEY=base64:BxTsckohrV35/sGUhb3UzNwUIJQj/02R8YfD7xyootY= \
-e TZ=Europe/London \
-v /home/containers/config/speedtest-tracker:/config \
--restart=always \
lscr.io/linuxserver/speedtest-tracker:latest

[ckbaxter]

I can confirm the issue, I tested with various settings and nothing work:
SPEEDTEST_PING_URL=1.1.1.1
SPEEDTEST_PING_URL=heise.de
SPEEDTEST_PING_URL=8.8.8.8

I switch back to v0.20.2…this version works fine for me