Etherion

in unraid Version 6.12.0-rc5 2023-05-01

this is so odd. I have two unraid servers one works with this and the other don't. We want to mirror docker hub as a pull through cache. The "manifest unknown" error is on the in the container that is the mirror docker registry. It happens when there is a pull request( i can see HTTP/1.1" 200 2 "" from clients) from one of the system(clients) on the network. here is the output of the add container page: ``` Pulling image: registry:2 IMAGE ID [2]: Pulling from library/registry. IMAGE ID [8a49fdb3b6a5]: Pulling fs layer. Downloading 100% of 3 MB. Verifying Checksum. Download complete. Extracting. Pull complete. IMAGE ID [58116d8bf569]: Pulling fs layer. Downloading 100% of 278 KB. Verifying Checksum. Download complete. Extracting. Pull complete. IMAGE ID [55f7fac5d815]: Pulling fs layer. Downloading 100% of 6 MB. Verifying Checksum. Download complete. Extracting. Pull complete. IMAGE ID [fe88d67f473c]: Pulling fs layer. Downloading 100% of 400 B. Verifying Checksum. Download complete. Extracting. Pull complete. IMAGE ID [32a19664cea7]: Pulling fs layer. Downloading 100% of 214 B. Verifying Checksum. Download complete. Extracting. Pull complete. Status: Downloaded newer image for registry:2 ``` TOTAL DATA PULLED: 9 MB ``` Command executiondocker run -d --name='DockerRegistry' --net='bridge' -e TZ="Europe/Budapest" -e HOST_OS="Unraid" -e HOST_HOSTNAME="Pac-Man-1" -e HOST_CONTAINERNAME="DockerRegistry" -l net.unraid.docker.managed=dockerman -l net.unraid.docker.icon='https://github.com/juusujanar/unraid-templates/raw/master/img/DockerRegistry-logo.png' -p '5000:5000/tcp' -v '/mnt/user/appdata/registry/config.yml':'/etc/docker/registry/config.yml':'rw' -v '/mnt/user/appdata/registry':'/var/lib/registry':'rw' 'registry:2' 4a2892e460026e1d8c8f724c2a30334afd8461a889ae8b5efac3f2ac3131e0a5 ``` The command finished successfully! I also put this config.yml to make them act as a ``` /mnt/user/appdata/registry/config.ym version: 0.1 log: fields: service: registry storage: cache: blobdescriptor: inmemory filesystem: rootdirectory: /var/lib/registry http: addr: :5000 headers: X-Content-Type-Options: [nosniff] health: storagedriver: enabled: true interval: 10s threshold: 3 proxy: remoteurl: https://registry-1.docker.io ``` I have tried everyting i can think of. stopping docker. deleting the docker.img, clearing all shared data. removing registry:2 image. no matter what I do it just don't want get the manifest from https://registry-1.docker.io i can do a pull request int eh cli. then the docker service on the unraid system gets the access its just that the DockerRegistry that is not.

I have notice that my /mnt/user/appdata/registry/ folder is not increasing in size anymore. So I took a look and found that the DockerRegistry mirror is having some kind of problem getting the manifest from docker hub. I delete everything and then start from a keen system and it having the same problem here is the logs form the docker container: docker logs -f DockerRegistry 192.168.47.141 - - [09/May/2023:11:37:42 +0000] "HEAD /v2/runpod/pytorch/manifests/latest HTTP/1.1" 404 96 "" "docker/23.0.5 go/go1.19.8 git-commit/94d3ad6 kernel/5.4.0-148-generic os/linux arch/amd64 UpstreamClient(Go-http-client/1.1)" time="2023-05-09T11:37:42.687812344Z" level=error msg="response completed with error" err.code="manifest unknown" err.detail="unknown tag=latest" err.message="manifest unknown" go.version=go1.16.15 http.request.host="192.168.47.100:5000" http.request.id=6bfb3a39-0a12-459a-824a-fdeedcf4608f http.request.method=HEAD http.request.remoteaddr="192.168.47.141:51660" http.request.uri="/v2/runpod/pytorch/manifests/latest" http.request.useragent="docker/23.0.5 go/go1.19.8 git-commit/94d3ad6 kernel/5.4.0-148-generic os/linux arch/amd64 UpstreamClient(Go-http-client/1.1)" http.response.contenttype="application/json; charset=utf-8" http.response.duration=4.856565ms http.response.status=404 http.response.written=96 vars.name="runpod/pytorch" vars.reference=latest using docker on the unraid system works. i can pull the images but the DockerRegistry can't. i have tried login or not. also tested with known public images. Configuration /mnt/user/appdata/registry This is where Registry will store it's data. Port 5000 Default connection port any idea where I am going wrong?

I have notice that my /mnt/user/appdata/registry/ folder is not increasing in size anymore. So I took a look and found that the DockerRegistry mirror is having some kind of problem getting the manifest from docker hub. I delete everything and then start from a keen system and it having the same problem here is the logs form the docker container: docker logs -f DockerRegistry 192.168.47.141 - - [09/May/2023:11:37:42 +0000] "HEAD /v2/runpod/pytorch/manifests/latest HTTP/1.1" 404 96 "" "docker/23.0.5 go/go1.19.8 git-commit/94d3ad6 kernel/5.4.0-148-generic os/linux arch/amd64 UpstreamClient(Go-http-client/1.1)" time="2023-05-09T11:37:42.687812344Z" level=error msg="response completed with error" err.code="manifest unknown" err.detail="unknown tag=latest" err.message="manifest unknown" go.version=go1.16.15 http.request.host="192.168.47.100:5000" http.request.id=6bfb3a39-0a12-459a-824a-fdeedcf4608f http.request.method=HEAD http.request.remoteaddr="192.168.47.141:51660" http.request.uri="/v2/runpod/pytorch/manifests/latest" http.request.useragent="docker/23.0.5 go/go1.19.8 git-commit/94d3ad6 kernel/5.4.0-148-generic os/linux arch/amd64 UpstreamClient(Go-http-client/1.1)" http.response.contenttype="application/json; charset=utf-8" http.response.duration=4.856565ms http.response.status=404 http.response.written=96 vars.name="runpod/pytorch" vars.reference=latest using docker on the unraid system works. i can pull the images but the DockerRegistry can't. i have tried login or not. also tested with known public images. Configuration /mnt/user/appdata/registry This is where Registry will store it's data. Port 5000 Default connection port any idea where I am going wrong?

I vote for this.

I have search this forum but nothing I read seems to explain what the reason for the "slow" transfers We have a unraid server running 6.10.3 with 3GBytes/s SSD RAID. The network speeds is 10Gbps but when we transfer a 0.5TB file it hovers at 150MiB/s Iperf3 shows that between the unraid server and the Ubuntu 20.04 client system there is 4.33 Gbits/sec. if I use DD to test its dopers down to 150MiB/s. Its the same if I copy a file. have tired ssh and mounting a folder with cifs on the client server and it just copies to that folder. both system have fast AMD epyc CPU's with 500GB ram. Both the unraid nas and the client system has RAID SSD's. tested +2893 MB/s I have no idea why we are seeing a 1/3 what we would expect any suggestions?