Increasing retry_count for cloudinit on windows

I’ve got a server I’m trying to deploy windows onto with MAAS, and by the looks of the logs, the network isn’t ready by the time cloud init comes round to actually running. If I manually start the service, it works (networking being fully up at that time)

From the docs, if I increase retry_count (and maybe retry_count_interval) in the cloudinit config file, on the server, then it should work.

However, I’m not finding anywhere to do this? and when I try using curtin’s late_commands with an in-target, I can’t echo (I’m guessing because it’s mounting a windows file system, which then has no commands)

anyone have any suggestions on what I can do with this?

I’m assuming it’s a generated file as part of the deployment, because it’s got a timestamp around when the deployment happens.

(maas 2.9 snap)

Hi @fuzzysteve, it may be worth trying to copy /var/snap/current/preseeds/curtin_userdata_windows.sample (to a file without .sample and editing that with the desired retry_count

It may also help to know the MAAS windows images use cloudbase-init rather than cloud-init, some parameters may differ https://cloudbase.it/cloudbase-init/.