[Bug]: Docker-Compose, LogConsumer & service restart

[mcornaton]

Module

Core

Testcontainers version

1.17.3

Using the latest Testcontainers version?

Yes

Host OS

Linux

Host Arch

x86

Docker version

Client: Docker Engine - Community
 Version:           20.10.18
 API version:       1.41
 Go version:        go1.18.6
 Git commit:        b40c2f6
 Built:             Thu Sep  8 23:11:43 2022
 OS/Arch:           linux/amd64
 Context:           default
 Experimental:      true

Server: Docker Engine - Community
 Engine:
  Version:          20.10.18
  API version:      1.41 (minimum version 1.12)
  Go version:       go1.18.6
  Git commit:       e42327a
  Built:            Thu Sep  8 23:09:30 2022
  OS/Arch:          linux/amd64
  Experimental:     false
 containerd:
  Version:          1.6.8
  GitCommit:        9cd3357b7fd7218e4aec3eae239db1f68a5a6ec6
 runc:
  Version:          1.1.4
  GitCommit:        v1.1.4-0-g5fd4c4d
 docker-init:
  Version:          0.19.0
  GitCommit:        de40ad0

What happened?

I use TestContainers to start Docker Compose and follow logs of services to files. When a service crashes and automatically restarts - which happens a lot on start-up as our test stack is not yet stable -, logs are not followed anymore. Here's a snippet of my code:

DockerComposeContainer<?> environment = new DockerComposeContainer<>(name + "_", composeFiles);
serviceNames.forEach(service -> environment.withLogConsumer(service, outputFrame -> writeLog(service, outputFrame.getBytes())));

whereas:

• name is a simple String, representing the role of the Docker Compose.
• composeFiles is a list of File pointing to our Docker Compose files - usually, we use a single file there.
• serviceNames is a list of String, representing the services declared in our Docker Compose files we'd like to log output.
• writeLog is a simple utility method that writes logs in some accessible file, depending on the service name.

Is there something I could do to mitigate the issue - to keep getting logs followed to my files?

---

Note: I can't simply dump docker logs <service> at the end of my tests, as I follow logs to achieve 2 goals:

1. If a test crashes, then I still get the full logs of all my services.
2. I have to change output files during execution, so getting logs line by line makes it easy to do.

Relevant log output

No response

Additional Information

No response

[kiview]

Thanks for the issue report. So your set of services is in general resilient and will eventually reach the correct running state, correct?

I would recommend configuring a set of WaitStrategies that allow the DockerComposeContainer to reach the stable state before attaching the LogConsumer. I am quite confident that this will work and make the issue irrelevant.

[mcornaton]

Thanks for the quick reply! That would be a solution for my personal case. However, I think this case could be extended. As an example, what if people want to test the behavior of a service when it is restarted? They should get all the logs which I think is hard to do as of now.

[paubor]

That is exactly the use case we are looking for. Grab a service from docker-compose and restart it, or even better: stop it, do some checks, start it again.