Description of problem: RHV needs to support guest-agent for CoreOS. Introduced here: https://bugzilla.redhat.com/show_bug.cgi?id=1764804
Why qemu-guest-agent matters for RHV? Because it is needed for successful OCP IPI deployment on RHV. For instance - troubleshooting is challenging, since we cannot collect logs: bz#1810438 due missing Guest IP address. Other management tasks that GA provides to RHV for the guest management. Why we need qemu-GA in container? Because it is the only way to install anything in RHCOS, and RHCOS instances are running in the VMs providing OCP on RHV. Is qemu-GA relevant to any other product in our portfolio today? OSP? If OSP has OCP IPI, it may. -> I do not know. CNV? Only if we start supporting RHCOS as a standard OS in a VM, outside of OCP use case. (or OCP on top of OCP?) -> Not today.
1. Deploying qemu-guest-agent requires to set the details of the registry (e.g., authentication, image location) and of the qemu-guest-agent service 2. Ideally, qemu-guest-agent needs to be deployed before the node joins the OCP cluster (e.g., to be able to gather logs during the deployment) 3. We'd also like to deploy qemu-guest-agent on the bootstrap VM (that doesn't join the OCP cluster) 4. For this, we need to set the data for deploying qemu-guest-agent through ignition 5. After the node joins the OCP cluster, it makes sense for MCO to control the lifecycle of the qemu-guest-agent container It makes little to no sense to let RHV adding the pieces that are required for deploying the qemu-guest-agent to the ignition script. It makes more sense to have the OCP installer configuring that and let the ignition content propagate through RHV.
due to capacity constraints we will be revisiting this bug in the upcoming sprint
still on track to be included in 4.7, waiting info from CoreOS team about including the QGA within CoreOS.
tested working with release openshift-install-linux-4.7.0-0.nightly-2020-12-04-013308 and oVirt 4.4.2.6 the IP address is reported back to oVirt without any manual intervention.
Verified on: openshift-4.7.0-0.nightly-2020-12-04-013308 Steps: 1. Deployed a OCP Cluster 2. Check on RHV the IPs from the VMs, once the installation was done 3. Logged in one of the vms and checked the status of guest-agent daemon # systemctl status qemu-guest-agent Results: ip showed on RHV UI as expected and guest agent daemon installed and running
did you notice which IPs were reported to the oVirt Engine ? ( external only or internal also )
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory (Moderate: OpenShift Container Platform 4.7.0 security, bug fix, and enhancement update), and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHSA-2020:5633