"authentication required" from docker build when running fly launch

Questions / Help Build debugging
1

Hi, I’m running fly launch on my Phoenix project using all the fly defaults in my docker file, but it’s failing with the following error:

Error error building: failed to solve with frontend dockerfile.v0: failed to create LLB definition: failed to do request: Head "https://registry-1.docker.io/v2/hexpm/elixir/manifests/1.13.3-erlang-22.3.4.10-debian-bullseye-20210902": dial tcp 54.210.12.153:443: i/o timeout

The BUILDER_IMAGE I’m using in the Dockerfile is ARG BUILDER_IMAGE="hexpm/elixir:1.13.3-erlang-22.3.4.10-debian-bullseye-20210902", which is the default fly gives me. When I click the link provided by the error, it shows me this json:

{"errors":[{"code":"UNAUTHORIZED","message":"authentication required","detail":[{"Type":"repository","Class":"","Name":"hexpm/elixir","Action":"pull"}]}]}

I’ve tried running fly auth login, and a few different build image s, but I always get the same error. Let me know if you have any ideas!

2

Update: I tried running with LOG_LEVEL=debug, and I now see:

Error error building: failed to solve with frontend dockerfile.v0: failed to read downloaded context: failed to load cache key: write /data/docker/overlay2/mpjhtxip3idboo153r0pg4ob2/diff/context: no space left on device
3

It looks like your remote builder ran out of disk space. This is a known issue we’re working on, you can get around by by running:

fly apps destroy <name-of-builder>

Your deploy log should show you the builder name, it’s something like fly-builder-*.

The original io timeout may or may not have been related.

4

Thanks, but still can’t get it to work. I got further this time, but now I get:

DEBUG <-- 200 https://api.fly.io/graphql (369.33ms) {"data":{"app":{"deploymentStatus":{"id":"a2654071-6a2a-78c1-de07-2ed2d98c76db","inProgress":false,"status":"failed","successful":false,"description":"Failed due to unhealthy allocations - no stable job version to auto revert to","version":0,"desiredCount":1,"placedCount":1,"healthyCount":0,"unhealthyCount":1,"allocations":[{"id":"d92ec00f-2a98-1e6e-0f27-af55a92f87d5","idShort":"d92ec00f","status":"running","region":"ewr","desiredStatus":"run","version":0,"healthy":false,"failed":false,"canary":false,"restarts":2,"checks":[{"status":"warning","serviceName":"tcp-8080"}]}]}}}}
***v0 failed - Failed due to unhealthy allocations - no stable job version to auto revert to and deploying as v1
5

That means either the process is crashing, or the health checks are failing.

If you run fly status --all you’ll get a list of running or failed instances. fly vm status <id> will give you more details about health checks and exit codes.

If this is a Phoenix app that you created before version 1.6.4, you’ll need to make sure you’ve changed a bunch of config files to get things listening on the right IPs and configured to connect to Postgres. That’s all documented here: Deploy an Elixir Phoenix Application (pre v1.6.3)

7

Thanks @kurt, that solved the issue for me. I have a different issue now but I’ll post that in a separate thread. Thanks!