[e2e] network bind to HostIP: unexpected source IP #18269
Labels
flakes
Flakes from Continuous Integration
kind/bug
Categorizes issue or PR as related to a bug.
locked - please file new issue/PR
Assist humans wanting to comment on an old issue or PR with locked comments.
Infrequent, but persistent enough to merit filing:
podman and podman-remote. Only root, never rootless. f37 and debian. It appears to be a consistent failure within a CI run (if it fails once, it fails on all ginkgo retries) but please grain-of-salt this: my flake logger only gets single-flakes under rare conditions. It does appear reasonable that once the surprise interface appears, it will linger.
The source address is always
10.128.0.x
. On Fedora, the string10.128.0
does not appear anywhere else in the logs, only in the flakes. On Debian, though, the full (same) source IP appears in various tests thatcat /etc/hosts
:I think instrumenting the e2e tests to run
ip a
(on the host) might offer troubleshooting hints on future failures. If I can't find a fix or workaround for this flake, I'll look into doing that.Podman run networking [It] podman run network bind to HostIP
The text was updated successfully, but these errors were encountered: