since the update (caching yes / no) docker-container crashes about dns HOT 13 CLOSED

I am testing it on my machine and it seems to work, would you mind sharing your docker-compose or docker run command?

Also why do you run it in the host network?

Also try disabling the block lists as you may be running out of memory, hence your OS killing the container. You could also try leaving the caching to on, this is only useful to turn off if you use i. e. pihole on top of unbound.

from dns.

I use the gui of the Docker-Application from Synology-NAS (no separately docker-compose nor docker-command) to create the container. I use the host network, because it was the only way to adress this container from my Pi-hole which runs in a independent container with mavlan-network. I don´t know what's now different to my past setup when the container was running correctly. The network bridge-mode also does not solve the problem. I also tried several DNS-providers.

Sorry for the inconvenience...

from dns.

No problem, did you try disabling some features as mentioned above? Set some environment variables to off such as CACHING or BLOCK_MALICIOUS to see if it works. Also, does it crash at start or after?

from dns.

I tryed it with disabled features as recommended. Nevertheless, the container crashes immediately with the start an immediately after the automatic restart.

from dns.

It seems the Unbound process is launched successfully and then killed by an outside thing, likely your host OS. I tested it on my Synology NAS as well and it works.

Did you check your memory usage?

from dns.

Hi Quentin, thank you for your tests and clarifications. I tried several things (try and error) but I can not find the problem. The memory usage is ok, also the cpu-utilization. I will search further. The problem definitely seems to be on my side.

from dns.

maybe try running it as root, by adding the option --user="root" (try using the command line if it's not available in the Synology UI)

from dns.

Running "as root" did not solve my problem. I can not find the error in my system. I think you can close this thread. I'll let you know if I know more ... thank you again

from dns.

Oh wait, I might know, what's your CPU architecture? Or at least what's your Synology model?

As Unbound is now built from source in the docker build and not simply downloaded and installed, that might be a problem.

from dns.

What you can also try with a terminal is:

wget https://github.com/qdm12/cloudflare-dns-server/archive/02be30420dd36a001a4889d0ab39101060a7af4c.zip -O dns.zip
unzip dns.zip
rm dns.zip
cd dns
docker build -t qmcgaw/cloudflare-dns-server .

to build an older version of the docker image. I'll add more documentation on how to do it, so that you can try with other commits to see when it started failing.

from dns.

I have a Synology DS 716+ with Intel Celeron N3150, Quad Core | Threads=4 | FPU=Yes | Package Arch=Braswell | RAM=DDR3 8 GB (own RAM-Upgrade).
I will try your suggestion ...
Thx

from dns.

Hi, I've added documentation here, please try it out 👍

EDIT: Your CPU is 64 bit (amd64 architecture), so not the problem 😞 I thought it could be an ARM architecture for example.

from dns.

It seems like the container crashes on my side with a Segmentation fault, I am not sure if that's a hardware problem or not really, as I have tried with various Unbound versions. For now, use the --restart=always flag or restart: always docker-compose config to remedy this.

I will close the issue for now. I am also reworking hard the whole image to run supervised by a Go written program and not shell, so that might solve your problem.

There is also a new Slack channel if you want to join to keep updated and participate in the discussion 👍

from dns.