nutanix / docker-machine

Rancher Node Driver for Nutanix AHV
https://www.nutanix.com/products/acropolis/virtualization
Mozilla Public License 2.0
13 stars 17 forks source link

Since Rancher 2.8.3 with Machine image "v0.15.0-rancher110" + latest centos rhel stream 9 Cluster nodes are looping on creating/deleting server #101

Closed med-jordan closed 6 months ago

med-jordan commented 7 months ago

Nutanix Cluster Information

AOS 6.5.5.5 pc.2022.6.0.9

Rancher Version

2.8.3

Rancher Node Driver Version

3.6

Node template

Debug Output

Expected Behavior

Successfull Node deployment without errors.

Actual Behavior

Since Rancher 2.8.3 with Machine image version "v0.15.0-rancher110" Cluster nodes are looping on creating/deleting server on new and existing clusters using centos Rhel stream 9 latest image. "command: sudo hostname testcl01-pool1-a9f476fd-slm2c && echo "testcl01-pool1-a9f476fd-slm2c" | sudo tee /etc/hostname err: inappropriate ioctl for device output: )"

Important Factors

References

tuxtof commented 6 months ago

Hello @med-jordan

I found the problem and it is unrelated to nutanix, in all case ssh connection is not managed by the rancher node driver but by the rancher-machine binary

In the rancher/machine:v0.15.0-rancher110 container they remove the ssh binary from the image (on purpose or not) so now the rancher-machine binary use a SSH Go native implementation as we can see in the log

SSH binary not found, using native Go implementation

And this is the reason why it broke against recent OS like centos-9-stream

tuxtof commented 6 months ago

fix by https://github.com/rancher/machine/pull/240