Erik Sipsma
Erik Sipsma
The `Connect` method served by a v2 runtime service is supposed to provide the Shim's PID and a given Task's PID. Right now, our host-side runtime shim just forwards the...
Right now when a container exits you get some confusing error logs like this: ``` time="2019-12-11T20:03:38.436612731Z" level=error msg="error copying io" ExecID= TaskID=container-80-3 error="read unix @->firecracker.vsock: use of closed network connection"...
In order to verify that [the quickstart guide](https://github.com/firecracker-microvm/firecracker-containerd/blob/master/docs/quickstart.md) works as expected, we should run the script as an e2e test in the CI system. This will likely require figuring out...
As detailed in this comment: https://github.com/firecracker-microvm/firecracker-containerd/pull/339#discussion_r352941338 There are some delicate dependencies between the stubdrive paths created by drive handler and where each jailer assumes the paths are located. One possible...
The Firecracker team currently runs Kata Container tests as part of their CI to ensure changes do not break them and expressed openness to something similar for Firecracker-containerd. We should...
It's often very helpful for debugging to have logs from the agent process running inside the VM. Getting those logs is not straightforward as by default they will stay inside...
@nmeyerhans brought up during a review that we currently we only do fairly limited validation of various paths provided to firecracker-containerd. For example, [the paths provided for the container rootfs](https://github.com/firecracker-microvm/firecracker-containerd/blob/38742eeb4fb5d549b31727a57b8c79e88b204b59/runtime/service.go#L775-L776)...
Currently, clients using the fc-control client are able to receive [grpc error codes](https://godoc.org/google.golang.org/grpc/codes) in error objects returned by API calls, but those are very generic error types that don't communicate...
Our current vsock dialer implementation does exponential backoff from 100ms to 1.6s before giving up. I encountered a situation in the real world in which this timeout was too short...
As mentioned in the comments on #160 we should consider whether we want to support the other OCI annotations that exist for grouping containers under a single shim or in...