You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I have spent a few hours trying to figure out why our config from 12 months ago stopped working.
We left the project aside for a little while and have since updated XCP-ng and XOA. It looked like there were changes in either or both made that caused the inline cloud-config to stop executing.
After setting up a local lab installation of XCP-ng and XOA I was able to find out that latest XOA could not set cloud-config on XCP-ng at all, but updating to the latest XCP-ng fixes this issue. I found a thread (which I unfortunately can't find back now that I am writing this bug report) that said to remove the XOA based hostname template "hostname: {name}%" as a diagnostic/workaround action.
I set the hostname of our cloud-config to a static hostname and with all components fully updated (XCP-ng, XOA, terraform-provider-xenorchestra) it started running the cloud-config again. I suspect there were changes made to xo-server to accomodate something relating to XCP-ng core components and that these changes need to be applied in the terraform provider as well.
Sorry I can't be more specific. The whole xen/xcp-ng/xoa/terraform stack is pretty involved and I am not very familiar with it's logging/debug functionality. I also don't have admin on the target cluster, so it's hard for me to get logs without setting up a local lab again. Hopefully this is enough information to pinpoint the issue or to ask the relevant people upstream.
Best regards,
John Simons
The text was updated successfully, but these errors were encountered:
Hi,
I have spent a few hours trying to figure out why our config from 12 months ago stopped working.
We left the project aside for a little while and have since updated XCP-ng and XOA. It looked like there were changes in either or both made that caused the inline cloud-config to stop executing.
After setting up a local lab installation of XCP-ng and XOA I was able to find out that latest XOA could not set cloud-config on XCP-ng at all, but updating to the latest XCP-ng fixes this issue. I found a thread (which I unfortunately can't find back now that I am writing this bug report) that said to remove the XOA based hostname template "hostname: {name}%" as a diagnostic/workaround action.
I set the hostname of our cloud-config to a static hostname and with all components fully updated (XCP-ng, XOA, terraform-provider-xenorchestra) it started running the cloud-config again. I suspect there were changes made to xo-server to accomodate something relating to XCP-ng core components and that these changes need to be applied in the terraform provider as well.
Sorry I can't be more specific. The whole xen/xcp-ng/xoa/terraform stack is pretty involved and I am not very familiar with it's logging/debug functionality. I also don't have admin on the target cluster, so it's hard for me to get logs without setting up a local lab again. Hopefully this is enough information to pinpoint the issue or to ask the relevant people upstream.
Best regards,
John Simons
The text was updated successfully, but these errors were encountered: