Joining cluster fails about rke2-aws-tf HOT 4 CLOSED

It's really hard to troubleshoot this. Sometimes there's 2 out fo 3 master nodes that are working, while the 3rd one does not have a kube-apiserver started (rke2-server service running) hence shown as OutOfService for the LB. And sometimes just one node is good while 2 are reporting the errors from the initial comment.

from rke2-aws-tf.

Hi @kaosmonk , this is unfortunately a known issue with rke2 and not specific to this implementation with terraform. You can see the full discussion, workarounds, and track the status here.

from rke2-aws-tf.

For me, I found that the /var/lib/cloud/instance/scripts/00_download.sh file that gets populated needs to be tweaked. The scenario is that awscli already exists and it is failing on the install.

[root@ip-10-2-5-28 ~]# /var/lib/cloud/instance/scripts/00_download.sh
Loaded plugins: amazon-id, search-disabled-repos
Package unzip-6.0-21.el7.x86_64 already installed and latest version
Nothing to do
  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100 33.5M  100 33.5M    0     0  47.0M      0 --:--:-- --:--:-- --:--:-- 47.0M
Found preexisting AWS CLI installation: /usr/local/aws-cli/v2/current. Please rerun install script with --update flag.
[root@ip-10-2-5-28 ~]# echo $?
1

I'll try and figure out how to modify the population of the /var/lib/cloud/instance/scripts/00_download.sh script but if anyone knows already please let me know. :)

from rke2-aws-tf.

Issues with preexisting awscli installations were resolved via this commit:

7b729db

from rke2-aws-tf.