feat(controller): offer QEMU the kernel datapath, gated per compute

Kernel eligibility becomes "both endpoints on one compute, each able to
anchor": Docker always (adapters are born as veth pairs), QEMU when that
compute's uBridge can create the persistent TAPs its adapters anchor on. The
question is asked per compute, because the link needs both ends anchored --
so a mixed or half-upgraded deployment degrades to the relay instead of
failing an attach.

probe_tap_support() creates and deletes one throwaway persistent TAP through a
scratch uBridge, cached by binary identity like the tc report; /capabilities
exposes `ubridge_tap` (absent = unknown, which keeps links on the relay).

The compute route for a QEMU adapter NIO took UDPNIO only, so a kernel link
was rejected with 422: its body and response now accept
Union[UDPNIO, BridgeNIO] exactly like Docker's.

Stale "Docker-to-Docker" wording in the config, link and MCP descriptions now
talks about anchors, and docs/features/qemu-kernel-datapath.md documents the
datapath -- verified end-to-end on a real instance: 32/32 kernel-datapath
checks and 29/29 on the relay variant.
This commit is contained in:
YueGuobin committed 2026-10-06 00:21:08 +08:00
1 parent 7d2d8c6747
commit 187201a08c
15 files changed
+444 -40

No files matched your search

+4 -3
View File
@@ -801,9 +801,10 @@ async def link_update(
Supported kwargs:
- suspend: boolean - Suspend or resume the link
- filters: dict - Packet filters (values MUST be positional arrays). Types
marked (K) run on kernel-datapath links only (Docker-to-Docker on one
compute) and are rejected with 409 on relay links; which types a given
link accepts is authoritatively answered by link_available_filters:
marked (K) run on kernel-datapath links only (both endpoints on one
compute with an anchor of their own) and are rejected with 409 on relay
links; which types a given link accepts is authoritatively answered by
link_available_filters:
* frequency_drop: [N] - Drop every Nth packet (N: -1 to 32767; -1 = drop
everything). Exact on kernel links (eBPF counter)
* packet_loss: [chance, correl?] - Loss percentage (0-100); correl (K)