Reputation: 61
CI job uses docker:dind image as a service for a specific build. However, in execution I get the error
.gitlab-ci.yml:
image: tmaier/docker-compose:latest
services:
- docker:18.09.7-dind
variables:
DOCKER_DRIVER: overlay2
stages:
- test
- build
- push
- clean-up
before_script:
- docker info
- apk update
- apk add make
- make down
test:
stage: test
tags: [test]
image: tiangolo/docker-with-compose:latest
script:
- make prod-up
- make composer-init-dev
- make test
config.toml:
concurrent = 1
check_interval = 0
[session_server]
session_timeout = 1800
[[runners]]
name = "My Docker Runner"
url = "https://###/"
token = "FRft8kukkz3qwXw6kkFJ"
executor = "docker"
pre_clone_script = "git config --global http.sslVerify false"
[runners.custom_build_dir]
[runners.docker]
tls_verify = false
image = "tiangolo/docker-with-compose"
privileged = true
disable_entrypoint_overwrite = false
oom_kill_disable = false
disable_cache = false
volumes = ["/cache"]
shm_size = 0
[runners.cache]
Insecure = false
[runners.cache.s3]
[runners.cache.gcs]
Error:
Server: ERROR: Cannot connect to the Docker daemon at tcp://docker:2375. Is the docker daemon running? errors pretty printing info ERROR: Job failed: exit code 1
All information about this problen https://gitlab.com/gitlab-org/gitlab-runner/-/issues/26727
Upvotes: 0
Views: 1371
Reputation: 5063
In case someone stumbles upon this question and doesn't want to read the comments and question edits to find the solution to the problem:
In OP's original CI configuration there were two docker-in-docker services declared:
docker:18.09.7-dind
docker:dind
It seems that they interfered with each other and when OP removed the job-specific one, everything started to work.
Upvotes: 1