diff --git a/issues/build/a-swaps-redial-races-a-hard-dial-ceiling-against-an-unbounded-guest-gap.md b/issues/build/a-swaps-redial-races-a-hard-dial-ceiling-against-an-unbounded-guest-gap.md
new file mode 100644
index 00000000000..0df0869b2aa
--- /dev/null
+++ b/issues/build/a-swaps-redial-races-a-hard-dial-ceiling-against-an-unbounded-guest-gap.md
@@ -0,0 +1,59 @@
+---
+status: expected-red
+kind: defect
+opened: 2026-09-28
+---
+
+# A swap's redial races a hard dial ceiling against an unbounded guest gap
+
+One mechanism, six sightings across `lan_swap`, `swap_netd` and
+`swap_crash_rolls_back`:
+
+- `lan_swap`, Fast tier.
+- `swap_netd` and `swap_crash_rolls_back`, Fast tier.
+- `lan_swap`, PR #535's nightly (run 36314576406, `guest (1)`, `a4f68c5a`, KVM, QEMU 11.1.0).
+- `swap_crash_rolls_back`, on origin/main's netd and on a branch's, under ten spinning host threads beside the run.
+- `swap_crash_rolls_back`, main's nightly (`1ce71831`, run 36290616312), not seen on the nightly before #527 (run 36285169430).
+- `swap_netd`, the rust-lld branch's fast tier (`e4317d3f`, PR #532), the one red of 400 besides `lan_mdns_answer`'s `SUN_LEN`.
+
+The guest side finished on every sighting that shows a console: today's three
+boots reached DHCP lease, `logd` back on port 41337, and init's own
+`restored`/`in service` line; `lan_swap`'s nightly guest reached `logd: serving
+this boot's log on port 41337` at 1.176 s and `init: swap netd: in service` at
+6.141 s, with no second `serving this boot's log to 10.0.2.2:…` line; both
+`swap_crash_rolls_back` sightings' consoles show the rollback completing
+(`restored`, then `logd` serving again). Only the host's redial gave up first,
+every time. Every sighting's `cargo run -- --known-red` answered NO (not
+quarantined), and an alone re-run is reliably green: 4 dials turned away on
+`lan_swap`'s nightly, `swap_netd` green in 10 s on the rust-lld branch,
+`swap_crash_rolls_back` green twice on main's nightly and once with netd
+reverted to origin/main.
+
+## What the code shows
+
+`metalswap::swap` arms `Stream::redial` after logd's `CARRIER_LEAVING` and the
+`go` (`src/metalswap.rs`), and `Stream::redial` opens a plain TCP dial to
+`logd`'s log-stream port (`toyos_logstream::CARRIER`). `serve`'s loop
+(`src/metaltalk.rs`) counts every dial that is refused, reset, or closes
+before a line, and redials **at once** — there is no wait between attempts,
+the comment names the refusal itself as the event — until a line arrives or
+`metalswap::TURNED_AWAY_CEILING` (64) is reached. So the redial spends a
+**fixed count** of dials against a gap whose length the *guest* sets: the old
+netd's exit, the new one's spawn and DHCP lease, and init's whole 5000 ms "in
+service" probe before it falls back to the one it replaced. This is
+`issues/diagnostics/a-swaps-redial-asks-again-with-no-event-to-wait-on.md`'s
+compromise.
+
+## What the measurement shows
+
+In `swap_netd`, 64 dials were refused in 326 ms between logd's `netd is being
+replaced` (3.061 s) and its re-listen (3.387 s), each taking 5 ms or less.
+Green swaps in the same runs were refused 3, 6 and 22 times. The dials got
+faster on the red runs, not slower, which points at the refusal window logd
+holds open while the old netd is still up.
+
+## Exit condition
+
+`Stream::redial` gives up on its time bound alone, never on a dial count, and a
+swap whose refusal window is staged long is green; the three rows come off with
+it. Owner: `src/metaltalk.rs`'s `Stream::redial`; held by the orchestrator.
diff --git a/issues/build/lan-swap-redial-spent-its-ceiling-on-a-nightly-shard.md b/issues/build/lan-swap-redial-spent-its-ceiling-on-a-nightly-shard.md
deleted file mode 100644
index e76e81fa8f1..00000000000
--- a/issues/build/lan-swap-redial-spent-its-ceiling-on-a-nightly-shard.md
+++ /dev/null
@@ -1,44 +0,0 @@
----
-status: open
-kind: defect
-opened: 2026-09-27
----
-
-# `lan_swap`'s redial spent its ceiling on a nightly shard
-
-PR #535's nightly (run 36314576406, `guest (1)`, a4f68c5a, KVM, QEMU 11.1.0):
-
-```
-FAIL lan_swap: 2 finding(s):
-  init's words on netd were ["accepted"] ending in None, where InService is owed (the stream had 1 connection(s) before the ask and 1 after)
-  the stream's redial was turned away 64 time(s), its ceiling of 64, and gave up
-```
-
-The guest completed the swap. Its console has `init: swap netd: in service`
-at 6.141 s, and `logd: serving this boot's log on port 41337` at 1.176 s
-through the new netd. After that `logd` admitted no reader: no second
-`serving this boot's log to 10.0.2.2:…` line. So none of the host's 64 dials
-reached `logd` once it listened again. That is consistent with all of them
-ending inside the guest's gap, which runs from `logd`'s `netd is being
-replaced` (1.000 s), through init stopping the old netd (1.080 s), to `logd`
-listening again (1.176 s). The alone re-run was green with 4 dials turned away.
-
-This is `issues/build/swap-crash-rolls-back-reds-when-its-redial-spends-its-ceiling-under-load.md`
-on the 82574 bench: the compromise
-`issues/diagnostics/a-swaps-redial-asks-again-with-no-event-to-wait-on.md`
-records, reached. A redial asks again at once, and the ceiling counts dials,
-not time. `lan_swap`'s path (`Ssh::swap`, `metalswap::swap`, `Stream::redial`,
-`logd`, netd, init) has no change on #535. Main's nightly at 16d2e645 (run
-36306830048) was green. Main's nightly at 1ce71831 (run 36290616312) had the
-same two findings on `swap_crash_rolls_back`.
-
-Dev host, QEMU 11.1.1, TCG, one named run each: `nightly-green2` at 877b8c95,
-`EXIT=0`, 12 dials turned away; `main` at 16d2e645, `EXIT=0`, 11. Not measured:
-how fast a KVM guest turns a dial away, and so how many dials fit inside the
-gap there.
-
-`cargo run -- --known-red lan_swap` answers NO.
-
-**Exit**: the redial waits on a guest-side event (the diagnostics issue's exit).
-Until then `lan_swap` reds on the nightly at a rate. It should go on #542's
-disabled list when that lands, citing this file.
diff --git a/issues/build/parallel-tests-red-under-other-suites.md b/issues/build/parallel-tests-red-under-other-suites.md
index df067417384..23580b7a723 100644
--- a/issues/build/parallel-tests-red-under-other-suites.md
+++ b/issues/build/parallel-tests-red-under-other-suites.md
@@ -505,11 +505,3 @@ mechanism for it.
   parallel run — a loaded full fast tier in which `i8042_undecoded_bytes`'
   first mute line names nothing and its second names the sequence, or the
   retirement's clause narrowed to the conditions under which it holds.
-
-- **`swap_netd`, again on its recorded signature.** The rust-lld branch's fast
-  tier at `e4317d3f` (PR #532), on a dev host whose load averages read 39.8,
-  31.1 and 37.5 as it started, from other worktrees: `the stream's redial was
-  turned away 64 time(s), its ceiling of 64, and gave up`, with init's words on
-  netd `["accepted"]` ending in `None`; the one red of 400 besides
-  `lan_mdns_answer`'s `SUN_LEN`, and `ALONE swap_netd: GREEN` in 10 s. The
-  branch touches neither netd, swap nor init. Not investigated here.
diff --git a/issues/build/swap-crash-rolls-back-redial-turned-away-once-on-mains-nightly.md b/issues/build/swap-crash-rolls-back-redial-turned-away-once-on-mains-nightly.md
deleted file mode 100644
index f9a3abde491..00000000000
--- a/issues/build/swap-crash-rolls-back-redial-turned-away-once-on-mains-nightly.md
+++ /dev/null
@@ -1,24 +0,0 @@
----
-status: open
-kind: finding
-opened: 2026-09-27
----
-
-# `swap_crash_rolls_back`'s redial was turned away to its ceiling once on main's nightly
-
-Main's nightly at 1ce71831 (run 36290616312), one guest shard, wide:
-
-```
-FAIL swap_crash_rolls_back: 2 finding(s):
-  init's words on netd were ["accepted"] ending in None, where Restored is owed (the stream had 1 connection(s) before the ask and 1 after)
-  the stream's redial was turned away 64 time(s), its ceiling of 64, and gave up
-```
-
-`ALONE swap_crash_rolls_back: GREEN` twice. It was not red on the nightly
-before #527 (run 36285169430).
-`cargo run -- --known-red swap_crash_rolls_back` answers NO.
-
-Not shown: what turned the redial away 64 times, and why.
-
-**Exit**: a cause for a redial turned away to its ceiling on a swap that
-rolled back, or a rate with enough runs to call it gone.
diff --git a/issues/build/swap-crash-rolls-back-reds-when-its-redial-spends-its-ceiling-under-load.md b/issues/build/swap-crash-rolls-back-reds-when-its-redial-spends-its-ceiling-under-load.md
deleted file mode 100644
index b5b806e18d2..00000000000
--- a/issues/build/swap-crash-rolls-back-reds-when-its-redial-spends-its-ceiling-under-load.md
+++ /dev/null
@@ -1,30 +0,0 @@
----
-status: open
-kind: defect
-opened: 2026-09-26
----
-
-# swap_crash_rolls_back reds when its redial spends its ceiling under host load
-
-`swap_crash_rolls_back` is red with the same two findings on origin/main's netd
-and on a branch's: the host stream's redial after the swap was turned away
-`metalswap::TURNED_AWAY_CEILING` (64) times and gave up, so init's `restored`
-never reached the host, although the guest's own console shows the rollback
-completing (`init: swap netd: restored`, then `logd: serving this boot's log on
-port 41337`). It is the compromise
-`issues/diagnostics/a-swaps-redial-asks-again-with-no-event-to-wait-on.md`
-records, reached: the redial asks again at once, and on a loaded host the
-ceiling runs out before `logd` listens again.
-
-Seen with ten spinning host threads beside the run, on the review branch of
-the netd receive-pipe fix: once in a full `-- swap` run of five, red again
-alone in that same run; and once in three `-- swap_crash_rolls_back` runs
-with netd reverted to origin/main (4b235d27), where the harness's alone re-run
-was green and called the `Sched::Parallel` classification wrong. The other
-five of those six runs, three on the branch's netd and two on main's, were
-green. `cargo run -- --known-red
-swap_crash_rolls_back` answers that it is not quarantined.
-
-Exit condition: the redial waits on a guest-side event (the linked issue's
-exit condition), or the test is shown green over a stated number of loaded
-runs.
diff --git a/issues/kernel/a-held-disk-waits-for-a-pass-no-cpu-takes-when-every-cpu-is-in-a-call-on-it.md b/issues/kernel/a-held-disk-waits-for-a-pass-no-cpu-takes-when-every-cpu-is-in-a-call-on-it.md
index d9bfc536b7b..df18b1914ba 100644
--- a/issues/kernel/a-held-disk-waits-for-a-pass-no-cpu-takes-when-every-cpu-is-in-a-call-on-it.md
+++ b/issues/kernel/a-held-disk-waits-for-a-pass-no-cpu-takes-when-every-cpu-is-in-a-call-on-it.md
@@ -1,5 +1,5 @@
 ---
-status: open
+status: expected-red
 kind: defect
 opened: 2026-09-22
 ---
@@ -43,3 +43,8 @@ the held disk; the interleaving comes at a rate.
 **Exit**: a disk call that waits for a device does not hold a CPU with `IF`
 clear — the wait parks — or a held call's CPU may bind within its bound without
 making the bind the caller's cost. Either is the owner's ruling to revisit.
+
+Related records — `usb_transport_break`'s three other open red modes, not this
+one: `issues/kernel/a-shutdown-on-a-held-usb-disk-left-a-cpu-deaf-to-a-tlb-shootdown.md`,
+`issues/build/usb-transport-break-flushedstick-can-break-after-the-reboot.md`, and
+`issues/boot-media/a-disk-whose-port-went-away-panics-the-boot-at-roots-hold.md`.
diff --git a/src/redlist.rs b/src/redlist.rs
index 9adcffd4c98..301519afafe 100644
--- a/src/redlist.rs
+++ b/src/redlist.rs
@@ -46,6 +46,10 @@ pub const DISABLED: &[Disabled] = &[
         issue: "issues/hardware/i8042-mouse-ends-four-packets-short-with-a-clean-exit.md",
     },
     Disabled { test: "kill_while_blocked", issue: "issues/kernel/deferred-release-outlives-its-syscall.md" },
+    Disabled {
+        test: "lan_swap",
+        issue: "issues/build/a-swaps-redial-races-a-hard-dial-ceiling-against-an-unbounded-guest-gap.md",
+    },
     Disabled { test: "latency_wake", issue: "issues/build/latency-wake-reds-on-the-dev-host-at-a-rate.md" },
     Disabled {
         test: "partition_claim_departure",
@@ -75,10 +79,22 @@ pub const DISABLED: &[Disabled] = &[
         test: "so_cache_refusals",
         issue: "issues/kernel/so-cache-refusals-saw-the-kernel-refuse-nothing-once.md",
     },
+    Disabled {
+        test: "swap_crash_rolls_back",
+        issue: "issues/build/a-swaps-redial-races-a-hard-dial-ceiling-against-an-unbounded-guest-gap.md",
+    },
+    Disabled {
+        test: "swap_netd",
+        issue: "issues/build/a-swaps-redial-races-a-hard-dial-ceiling-against-an-unbounded-guest-gap.md",
+    },
     Disabled {
         test: "usb_disk_index_stable",
         issue: "issues/hardware/usb-disk-index-stable-nothing-enumerates-on-the-first-controller.md",
     },
+    Disabled {
+        test: "usb_transport_break",
+        issue: "issues/kernel/a-held-disk-waits-for-a-pass-no-cpu-takes-when-every-cpu-is-in-a-call-on-it.md",
+    },
     Disabled {
         test: "xhci_flap",
         issue: "issues/hardware/a-collapsed-replug-is-enumerated-only-when-another-port-event-arrives.md",