-
Notifications
You must be signed in to change notification settings - Fork 161
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
High disk usage during docker:dind scenarios with Testcontainers #875
Comments
Hi @mawl, can you provide more info on the setup and how to reproduce the problem? I see you are installing Sysbox on a Rancher K8s cluster, but what Testcontainers workload are you running on it? And when does the problem manifest itself? Is it somehow related to the number (or size) or images created by the Docker-in-Docker engine? In other words, does the problem occur when the Docker engine running inside the Sysbox pod has too many images? And does it occur when the pod gets started or stopped, or while the pod is running? The more info you can provide, the better. Thanks. |
@ctalledo: Here are more insights - we're using a 1TB HD to catch high disk usage peaks. full list of volumes during this peak:
We have identified the cause, a dotnet test run in a GitLab CI job using testcontainers with NUnit, which runs: The developer has told me, that they are starting around 1-3 containers (Redis, MinIO, MongoDB) per test, for a test suite with 350 tests, which leads to 500-600 containers for a whole run.
It's more the number of containers, less too many different or big docker images.
It occurs while the pod/CI job is running. Remarkable: Their test suite had no issues before when running on a docker-executor GitLab Runner with privileged flag enabled. |
Hi @mawl, apologies for the belated response. Looking at the diagrams, I can see the disk usage in primarily in This suggests that either (1) many containers are running simultaneously or (2) containers are not being removed after stopping. Question: after installing sysbox via sysbox-deploy-k8s, do you experience the same problem when running the CI jobs without sysbox (i.e., using the default The other place I would look at is the CRI-O logs on the k8s node (e.g., Finally, the diagrams don't shows the storage under |
Hey,
We're seeing high disk usage and defective CI jobs on rancher worker nodes once or twice a week, mainly in docker-dind scenarios using Testcontainers. An update to Sysbox v0.6.5 doesn't fix this.
sudo journalctl -eu crio
Thanks for any help in advance.
The text was updated successfully, but these errors were encountered: