trunk: the allowed list keeps the four media defaults, so a remainder is 13-4094 #1946
No reviewers
Labels
No milestone
No project
No assignees
1 participant
Notifications
Due date
No due date set.
Dependencies
No dependencies set.
Reference
jaylouisw/packetforge!1946
Loading…
Add table
Add a link
Reference in a new issue
No description provided.
Delete branch "rev/1118-0501"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Carried over from the 05:01 reverse-engineering run, whose push never reached the server: the commit sat in the worktree while Forgejo answered
Internal Server Connection Error. It is pushed now, unmodified.switchport trunk allowed vlan remove <list>on a trunk that still allows the defaultallused to print1-4,7,11,13-1001,1006-4094- two runs around VLANs 1002-1005, which no box prints. A trunk's allowed list is 802.1Q tags over 1-4094 and the four media defaults are in it (Cisco's c3550 transcript ofremove 5-10,12+add 7prints1-4,7,11,13-4094), so they cannot be removed from it; the parse rule that keeps a port out of VLAN 1002 was being applied to a trunk's tag list.Detail and sources:
research/trunk-allowed-reserved-vlans.md.Tests:
trunk-allowed-reserved-vlans29 passed,trunk-allowed-remove-from-all45 passed; the full suite on this tree ran 334 suites / 11,588 assertions / 0 failed.Surgery loop: one leg red, not reproducible - needs a re-run, not a rebase.
Head
721fa5e1:Test (Node 18)success,Test (Node 22)success,Test (Node 20)failure. Everything else green,Arm auto-mergearmed.Classified as not superseded: merging
origin/main(5f1ce466) into this branch is clean - auto-mergedengine/cli.js, noCONFLICT (add/add), nothing already on main.Local reproduction attempt (Node v24.11.0), head merged into main:
So the trunk/vlan surface this PR touches is green here, and green on Node 18 and 22 in CI. With 18 and 22 both passing, a genuine Node-20-only code defect is unlikely - it looks like a flake or a Node-20 leg that hit the host while it was still recovering (the API was answering DB queries in ~70 s at 06:23 and the test legs on #1942 sat pending for 35+ minutes).
What I need: a re-run of the Node 20 leg (or a push that re-triggers it). I am not merging on a red head-SHA status - that is the standing rule - and I am not touching
engine/ortest/to chase a failure I cannot see. I also cannot fetch the job log: this Forgejo build has no/actions/runsor/actions/jobs/*/logsAPI, so the failure text is only visible in the web UI.If the re-run goes red again on Node 20 only, that is a real Node-20 issue and I will escalate it with the failing suite name once someone pastes the log.
721fa5e16d377ddbc688377ddbc68872b4e6dc1f