Closed ValHeimer closed 3 years ago
Does anyone know if it is indeed a slirp4netns issue or if I should open an issue somewhere else ? (@AkihiroSuda ?) Thank you!
It was an error in my host config........
I just had to run
sudo sysctl net.ipv4.conf.all.rp_filter=2
recreate the pod, and it works
sorry about that
repost of this issue https://github.com/containers/podman/issues/10923 as it seems to seem a slirp4netns bug
Is this a BUG REPORT or FEATURE REQUEST? (leave only one on its own line)
/kind bug
Description
In a rootless pod, the ip rule "uidrange" does not seem to work properly.
Steps to reproduce the issue:
In the pod, I have 2 containers:
Add routing rules (from inside the pod, in the container with user root)
add a rule AFTER the uidrange rule whish should not be catch in our test
ip rule add to 1.1.1.1/24 table 666
then add a route in table 666 which can't resolve 1.1.1.1, this is only to "nullroute" the request if matched
$
ip rule
$
id
Describe the results you received: The curl request to 1.1.1.1 times out (idk why, maybe because the routing rule didn't make the match with the uidrange for the response, it seems that the corresponding response is seen as user uid 0 (it works )
Describe the results you expected: The curl request should succeed (the corresponding response should be seen as user uid 50000)
Output of
podman version
:Output of
podman info --debug
:Package info (e.g. output of
rpm -q podman
orapt list podman
):Additional environment details (AWS, VirtualBox, physical, etc.):
VMware