No such image after downloading image

I’m using a private Docker image that CircleCI seems to successfully pull but cannot find afterwards. The credentials are fine (I get an authorization error when they’re not) and I’m able to successfully run locally using circleci local execute. Is there something I’m doing wrong or is this a service issue?

23302e52b49d: Pull complete
cf5693de4d3c: Pull complete
fd542ee25159: Pull complete
63abcb3c0012: Pull complete
458fad138554: Pull complete
bcfe8987f677: Pull complete
5a33f0f60b8a: Pull complete
1dbdebfe03b0: Pull complete
a50b3c4511a6: Pull complete
655dab78f7f7: Pull complete
882e8e32bbac: Pull complete
9ff6cea3a4b1: Pull complete
ac20b476753a: Pull complete
684460869e67: Pull complete
072c5b80aa38: Pull complete
1225463c4737: Pull complete
error starting container myorg/myimage@sha256:0db403d47bc6156221b3602889906f811bad105e42c3f29b115e70e0b2c43cfb: Error: No such image: myorg/myimage@sha256:0db403d47bc6156221b3602889906f811bad105e42c3f29b115e70e0b2c43cfb

I’m seeing this as well

I was able to fix the issue after removing an environment variable I had set as well as aws-cli. I suspect the culprit was the environment variable which was MY_REPO_NAME_REPO.

We were seeing this intermittently in workflows that use the same container for multiple simultaneous build jobs. It was failing transiently in random jobs where 9 out 10 jobs would pull the image and run just fine and one would fail. On the next build, a different job would fail with the same error and the one that failed in the last build would run fine.

(I don’t think your config was the problem, rather there was something weird going on with parts of the Circle Docker infrastructure)