You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
A side note that I noticed while making this : not all logs and errors that we are generating are actually getting displayed when using youki. For example, as of now, we verify the binary before running the container, and error if it does not exist. However, the podman log simply errors out, not showing anything, whereas with runc proper log of
missing executable is shown :
Youki
podman run --runtime $PWD/youki busybox /bin/bash
DEBUG youki: started by user 0 with ArgsOs { inner: ["...../youki/youki", "delete", "--force", "9c937d43d590695b71556382fd6d9f4baa51c9780c7969b94e7c7753ad542487"] }
DEBUG youki::commands::delete: start deleting 9c937d43d590695b71556382fd6d9f4baa51c9780c7969b94e7c7753ad542487
Error: container create failed (no logs from conmon): EOF
runc
podman run --runtime runc busybox /bin/bash
Error: runc create failed: unable to start container process: exec: "/bin/bash": stat /bin/bash: no such file or directory: OCI runtime attempted to invoke a command that was not found
I'm pretty sure that we are generating similar error and propagating that upwards. This is similarly observed in actual journalctl logs (after always enabling them for debug in this PR) , that not all of tracing logs are shown. cCan this be because with namespace change that we do in intermediate/init process, those logs are not getting forwarded properly?
The text was updated successfully, but these errors were encountered:
Summary:
A side note that I noticed while making this : not all logs and errors that we are generating are actually getting displayed when using youki. For example, as of now, we verify the binary before running the container, and error if it does not exist. However, the podman log simply errors out, not showing anything, whereas with runc proper log of
missing executable is shown :
Youki
podman run --runtime $PWD/youki busybox /bin/bash
DEBUG youki: started by user 0 with ArgsOs { inner: ["...../youki/youki", "delete", "--force", "9c937d43d590695b71556382fd6d9f4baa51c9780c7969b94e7c7753ad542487"] }
DEBUG youki::commands::delete: start deleting 9c937d43d590695b71556382fd6d9f4baa51c9780c7969b94e7c7753ad542487
Error: container create failed (no logs from conmon): EOF
runc
podman run --runtime runc busybox /bin/bash
Error: runc create failed: unable to start container process: exec: "/bin/bash": stat /bin/bash: no such file or directory: OCI runtime attempted to invoke a command that was not found
I'm pretty sure that we are generating similar error and propagating that upwards. This is similarly observed in actual journalctl logs (after always enabling them for debug in this PR) , that not all of tracing logs are shown. cCan this be because with namespace change that we do in intermediate/init process, those logs are not getting forwarded properly?
The text was updated successfully, but these errors were encountered: