taodeng/aci-sim
1
0
Fork 0
mirror of https://github.com/dtzp555-max/aci-sim.git synced 2026-07-19 09:46:33 +00:00
Code Issues Packages Projects Releases Wiki Activity

fix(isn): flip sub-interface naming to real-gear conventions (v0.19.1)

Browse Source 
Owner review of the v0.19.0 demo + Cisco docs/real spine CLI verification:
ACI spines name the VLAN-4 sub-if after the PORT (eth1/49.49, cf.
Eth1/29.29-style in show ip ospf neighbors vrf overlay-1); the ISN Nexus
names its dot1q-4 sub-if after the VLAN (Ethernet1/1.4). v0.19.0 had the
two conventions swapped. 908 tests green.
This commit is contained in:
dtzp555-max
2026-07-07 08:55:16 +10:00
parent f6c95a0105
commit c926702a29
9 changed files with 57 additions and 32 deletions
Download Patch File Download Diff File Expand all files Collapse all files
docs/CONTRACT.md
+1 -1
View File
@@ -312,7 +312,7 @@ target-subtree-class=lldpAdjEp`) returned `totalCount=0` because the root DN had
`--json` for machine-readable output).
PR-19: `ospfIf.area`/`ospfAdjEp.area` are now `isn.ospf_area`-driven (default `"0.0.0.0"`) instead of a hardcoded literal.
PR-20: `ospfIf.helloIntvl`/`deadIntvl` are new, driven by `isn.ospf_hello`/`isn.ospf_dead` (defaults 10/40, real ACI defaults) — previously not carried on this MO at all.
PR-22: the ISN-facing `ospfIf`/`ospfAdjEp` on each multi-site spine now sit on a routed VLAN-4 sub-interface (`id`/dn `eth1/{49+si}.4`, was the bare physical port `eth1/{49+si}`) and `ospfIf` gained an `addr` attribute (`172.16.{site.id}.{si+1}/24`) — matching real ACI multi-site spines, which carry the ISN OSPF address on a sub-if rather than the physical port. The underlying physical port's `l1PhysIf`/`ethpmPhysIf` (interfaces.py) are unchanged. Each spine's ISN uplink physical port also gained a simulated `lldpAdjEp`/`cdpAdjEp` (`sysName=ISN-CSW{site.id}`, `portIdV=Ethernet1/{si+1}`, model `N9K-C9364C`) via `build/fabric.py`, the third `add_adjacency` call site — see docs/DESIGN.md's "PR-22" section.
PR-22: the ISN-facing `ospfIf`/`ospfAdjEp` on each multi-site spine now sit on a routed VLAN-4 sub-interface (`id`/dn `eth1/{49+si}.{49+si}` — port-named per real ACI spine CLI (Eth1/29.29-style); was the bare physical port `eth1/{49+si}`) and `ospfIf` gained an `addr` attribute (`172.16.{site.id}.{si+1}/24`) — matching real ACI multi-site spines, which carry the ISN OSPF address on a sub-if rather than the physical port. The underlying physical port's `l1PhysIf`/`ethpmPhysIf` (interfaces.py) are unchanged. Each spine's ISN uplink physical port also gained a simulated `lldpAdjEp`/`cdpAdjEp` (`sysName=ISN-CSW{site.id}`, `portIdV=Ethernet1/{si+1}.4` (the ISN Nexus names its dot1q-4 sub-if after the VLAN), model `N9K-C9364C`) via `build/fabric.py`, the third `add_adjacency` call site — see docs/DESIGN.md's "PR-22" section.
**Overlay/BGP:** `bgpInst`(dn,asn) · `bgpPeer`(dn,asn,addr,type,peerRole) ·
`bgpPeerEntry`(dn,addr,operSt∈{established,...},rtrId,lastFlapTs,connEst,connDrop,type) ·
docs/DESIGN.md
+4 -4
View File
@@ -81,7 +81,7 @@ Each sub-builder is `build(topo, site, store)` and only ADDS MOs. `orchestrator.
Sub-builders (one concern each, <~200 lines):
- `fabric.py` — fabricNode (roles/ids/model/serial/version from YAML), topSystem, fabricHealthTotal, vpcDom/vpcIf (border-leaf pair), infraWiNode.
- `cabling.py` — fabricLink (with `/lnk-…/` DN), lldpAdjEp + cdpAdjEp derived STRICTLY from the cabling graph, isisAdjEp.
- `underlay.py` — bgpInst (local ASN), ospfAdjEp/**ospfIf**✅ (spine↔ISN, on the routed VLAN-4 sub-if `eth1/{49+si}.4` with `addr` — PR-22; multi-site only), IS-IS dom.
- `underlay.py` — bgpInst (local ASN), ospfAdjEp/**ospfIf**✅ (spine↔ISN, on the routed VLAN-4 sub-if `eth1/{49+si}.{49+si}` (port-named) with `addr` — PR-22; multi-site only), IS-IS dom.
- `overlay.py` — intra-site BGP-EVPN (spines as RR, leaves as clients): bgpPeer/bgpPeerEntry(established)/bgpPeerAfEntry;
**ISN**: on each spine's `sys/bgp/inst`, add inter-site bgpPeer with peer addr /32 + remote (other-site) ASN.
- `endpoints.py` — fvCEp/fvIp distributed across leaves+EPGs (encap/fabricPathDn consistent), epmMacEp/epmIpEp, fvIfConn.
@@ -835,15 +835,15 @@ ACI hardware is the authoritative backward-compat gate for this PR.
modeled the ISN-facing `ospfIf` as a **plain interface** (`eth1/{49+si}`) with **no address**, and there
was no LLDP neighbor on that port at all — so the demo UI showed `Port=eth1/49`, `Local IP=-`,
`Neighbor Port=-`. Real ACI multi-site spines instead use a **routed sub-interface** (VLAN-4 encap, e.g.
`eth1/49.4`) that carries the OSPF address, while the ISN switch advertises LLDP on the underlying
`eth1/49.49`, port-named per real spine CLI) that carries the OSPF address, while the ISN switch advertises LLDP on the underlying
physical port. This PR makes the sim match that reality.
### `build/underlay.py` — `ospfIf` becomes a VLAN-4 routed sub-interface
`ospfIf.id`/dn now read `eth1/{49+si}.4` (was `eth1/{49+si}`), and the MO gained a new `addr` attribute:
`ospfIf.id`/dn now read `eth1/{49+si}.{49+si}` (was `eth1/{49+si}`), and the MO gained a new `addr` attribute:
`172.16.{site.id}.{si+1}/24` — the same `/24` as the existing OSPF peer `172.16.{site.id}.254`, so the
adjacency stays subnet-consistent. `area`/`helloIntvl`/`deadIntvl`/`operSt` are unchanged. `ospfAdjEp`'s
dn moves under the new sub-if segment (`if-[eth1/{49+si}.4]` instead of `if-[eth1/{49+si}]`); its
dn moves under the new sub-if segment (`if-[eth1/{49+si}.{49+si}]` instead of `if-[eth1/{49+si}]`); its
`peerIp`/`nbrId`/`operSt`/`area` attrs are unchanged. The underlying physical port
(`l1PhysIf`/`ethpmPhysIf`, built by `build/interfaces.py`) is **untouched** — only the OSPF logical
interface becomes a sub-interface; LLDP/physical port state stays on the plain port, matching real