This message was deleted.
# rke2
a
This message was deleted.
g
https://docs.cloud-init.io/en/21.2/topics/examples.html According to the documentation:
Copy code
# Note: Don't write files to /tmp from cloud-init use /run/somedir instead.
 # Early boot environments can race systemd-tmpfiles-clean LP: #1707222.
So the solution is to use:
Copy code
"export TMPDIR=/run",
    "curl -sfL <https://get.rke2.io> | INSTALL_RKE2_CHANNEL=${local.rke2_channel} INSTALL_RKE2_TYPE=agent sh -",
I wonder if anyone faced the same problem
I am using Leap Micro
c
I’ve not run into that before. But it seems easy enough to just put that in with the rest of the env vars if that’s what you need to do…
curl -sfL <https://get.rke2.io> | TMPDIR=/run INSTALL_RKE2_CHANNEL=${local.rke2_channel} INSTALL_RKE2_TYPE=agent sh -
g
Of corse - thats equivalent of
export TMPDIR=/run
, both work. I've found this: https://bugs.launchpad.net/cloud-init/+bug/1707222 For me it's happening on Harvester HCI/Rancher/Leap Micro I've don't know yet if the fix works, I haven't encountered the problem since switching to /run, but it's a race condition, so you never know ...
Looks like it didn't help, I faced one problem with one node today as well.
"sleep 5", "curl -sfL https://get.rke2.io | INSTALL_RKE2_CHANNEL=${local.rke2_channel} INSTALL_RKE2_TYPE=agent sh -", Works, but ... no need to explain ...