Hi there,
I’m currently troubleshooting nested virtualization across multiple guest OS environments (Linux, Windows, etc.) and I’m seeing consistent issues that seem to stem from IOMMU or PCI passthrough behavior, regardless of the guest OS.
I'm using an AMD Ryzen Threadripper 2920X (Family 17h), and I’m preparing to escalate this issue with a paid support ticket, but before doing so, I’d appreciate some clarification on how support tickets are handled.
I’m consistently encountering problems with nested virtualization environments, specifically around passthrough and IOMMU-related behavior. These issues seem to affect all guest operating systems in the same way, and may be tied to the CPU architecture or how PCI devices are exposed in the virtual hierarchy.
Here’s some relevant info from my system:
CPU Details (lscpu):
PCI Topology (lspci — sample):
I’d like to know if this issue would be covered under one paid support ticket, and what level of help can be expected. I’d also appreciate any guidance on what else I should include to make a potential ticket smoother for your team.
Thanks in advance for any help or clarification!
— Best Regards
I’m currently troubleshooting nested virtualization across multiple guest OS environments (Linux, Windows, etc.) and I’m seeing consistent issues that seem to stem from IOMMU or PCI passthrough behavior, regardless of the guest OS.
I'm using an AMD Ryzen Threadripper 2920X (Family 17h), and I’m preparing to escalate this issue with a paid support ticket, but before doing so, I’d appreciate some clarification on how support tickets are handled.
Questions:
- Is a paid support ticket charged per question or per issue?
For instance, if I’m debugging this virtualization problem across multiple guests and hypervisors, would that be considered one issue? - What is the scope of support for a paid ticket?
Does it include:
- Hardware-level diagnostics?
- PCIe topology/IOMMU behavior in nested virtualization setups?
- Troubleshooting across hypervisors (e.g., KVM, VMware, Proxmox)?
- What information should I prepare before opening a ticket?
Are lspci, lscpu, and kernel logs enough to get started, or is there a preferred diagnostic format?
Issue Context:
I’m consistently encountering problems with nested virtualization environments, specifically around passthrough and IOMMU-related behavior. These issues seem to affect all guest operating systems in the same way, and may be tied to the CPU architecture or how PCI devices are exposed in the virtual hierarchy.
Here’s some relevant info from my system:
CPU Details (lscpu):
Code:
Architecture: x86_64
Model name: AMD Ryzen Threadripper 2920X 12-Core Processor
Socket(s): 1
Core(s) per socket: 12
Thread(s) per core: 2
CPU(s): 24
Virtualization: AMD-V
NUMA node(s): 1
NUMA node0 CPU(s): 0-23
Flags: svm, npt, sev, sev_es, etc.PCI Topology (lspci — sample):
Code:
00:00.0 Host bridge: AMD Family 17h Root Complex
00:00.2 IOMMU: AMD Family 17h I/O Memory Management Unit
00:01.1 PCI bridge: AMD PCIe GPP Bridge
00:14.0 SMBus: AMD FCH SMBus Controller
08:00.0 VGA controller: NVIDIA RTX 3090
08:00.1 Audio device: NVIDIA HD Audio Controller
09:00.2 Encryption: AMD PSP 3.0 Device
0a:00.2 SATA controller: AMD FCH SATA Controller (AHCI)
41:00.0 NVMe: Intel SSD 660P SeriesSystem Summary:
- CPU: AMD Ryzen Threadripper 2920X (Family 17h)
- Board/Chipset: AMD X399 Series
- Hypervisors tested: KVM/QEMU, VMware, Proxmox
- Issue: Nested guest OSes (Linux/Windows) fail or behave inconsistently when IOMMU or device passthrough is used.
- Consistent behavior across guests: Yes
- Suspected cause: Possibly IOMMU group exposure or AMD-V features not behaving as expected in nested environments.
Request:
I’d like to know if this issue would be covered under one paid support ticket, and what level of help can be expected. I’d also appreciate any guidance on what else I should include to make a potential ticket smoother for your team.
Thanks in advance for any help or clarification!
— Best Regards
