pytest: fix reconnect flake in test_route_by_old_scid
What the developer wrote
Authored by Rusty Russell
pytest: fix reconnect flake in test_route_by_old_scid
We restart l2, then try to connect to l1. But l1 will be trying to
reconnect (and it can, since it was initially given l2's address), so
it can race us and we end up disconnecting because of simultaneous
connect:
```
2026-01-13T04:06:10.7490260Z > l2.rpc.connect(l1.info['id'], 'localhost', l1.port)
2026-01-13T04:06:10.7490667Z
2026-01-13T04:06:10.7490828Z tests/test_splicing.py:554:
...
2026-01-13T04:06:10.7525780Z > raise RpcError(method, payload, resp['error'])
2026-01-13T04:06:10.7527464Z E pyln.client.lightning.RpcError: RPC call failed: method: connect, payload: {'id': '0266e4598d1d3c415f572a8488830b60f7e744ed9235eb0b1ba93283b315c03518', 'host': 'localhost', 'port': 46813}, error: {'code': 402, 'message': 'disconnected during connection'}
2026-01-13T04:06:10.7528832Z
...
2026-01-13T04:06:10.9411680Z lightningd-2 2026-01-13T03:59:59.463Z DEBUG 0266e4598d1d3c415f572a8488830b60f7e744ed9235eb0b1ba93283b315c03518-connectd: Connect OUT
2026-01-13T04:06:10.9412178Z lightningd-2 2026-01-13T03:59:59.463Z DEBUG 0266e4598d1d3c415f572a8488830b60f7e744ed9235eb0b1ba93283b315c03518-connectd: peer_out WIRE_INIT
2026-01-13T04:06:10.9412551Z lightningd-2 2026-01-13T03:59:59.463Z DEBUG 0266e4598d1d3c415f572a8488830b60f7e744ed9235eb0b1ba93283b315c03518-connectd: Connect IN
2026-01-13T04:06:10.9412947Z lightningd-2 2026-01-13T03:59:59.464Z DEBUG 0266e4598d1d3c415f572a8488830b60f7e744ed9235eb0b1ba93283b315c03518-connectd: peer_out WIRE_INIT
2026-01-13T04:06:10.9413342Z lightningd-2 2026-01-13T03:59:59.464Z DEBUG 0266e4598d1d3c415f572a8488830b60f7e744ed9235eb0b1ba93283b315c03518-connectd: peer_in WIRE_INIT
2026-01-13T04:06:10.9413624Z lightningd-2 2026-01-13T03:59:59.487Z TRACE lightningd: Calling peer_connected hook of plugin chanbackup
2026-01-13T04:06:10.9414057Z lightningd-2 2026-01-13T03:59:59.534Z DEBUG 0266e4598d1d3c415f572a8488830b60f7e744ed9235eb0b1ba93283b315c03518-connectd: Handed peer, entering loop
2026-01-13T04:06:10.9414451Z lightningd-2 2026-01-13T03:59:59.534Z DEBUG 0266e4598d1d3c415f572a8488830b60f7e744ed9235eb0b1ba93283b315c03518-connectd: peer_in WIRE_INIT
2026-01-13T04:06:10.9414851Z lightningd-2 2026-01-13T03:59:59.574Z DEBUG 0266e4598d1d3c415f572a8488830b60f7e744ed9235eb0b1ba93283b315c03518-lightningd: peer reconnected
2026-01-13T04:06:10.9415250Z lightningd-2 2026-01-13T03:59:59.609Z DEBUG 0266e4598d1d3c415f572a8488830b60f7e744ed9235eb0b1ba93283b315c03518-lightningd: peer_disconnected
2026-01-13T04:06:10.9415527Z lightningd-2 2026-01-13T03:59:59.685Z TRACE lightningd: Calling peer_connected hook of plugin chanbackup
2026-01-13T04:06:10.9415957Z lightningd-2 2026-01-13T03:59:59.687Z DEBUG 0266e4598d1d3c415f572a8488830b60f7e744ed9235eb0b1ba93283b315c03518-connectd: Handed peer, entering loop
2026-01-13T04:06:10.9416364Z lightningd-2 2026-01-13T03:59:59.757Z DEBUG 0266e4598d1d3c415f572a8488830b60f7e744ed9235eb0b1ba93283b315c03518-lightningd: peer_disconnected
2026-01-13T04:06:10.9417067Z lightningd-2 2026-01-13T03:59:59.830Z DEBUG 0266e4598d1d3c415f572a8488830b60f7e744ed9235eb0b1ba93283b315c03518-connectd: Will try reconnect in 4 seconds
```
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
What changed, and why it matters
This is a one-line change to a test file that fixes an occasional test failure (a 'flake'). It swaps which test node initiates a connection after a restart to avoid a race where both nodes try to connect to each other at the same time. It does not change any production code or fix a security bug.
No security action needed. This is a test-only reliability fix.
Security signals we found
No strong security signals were identified.
Evidence from the diff
The commit modifies tests/test_splicing.py in the Core Lightning repository. In test_route_by_old_scid, after restarting l2, the test previously had l2 call connect() to l1. Because l1 was already configured with l2’s address and would attempt to reconnect automatically, the two directions could collide, causing a ‘disconnected during connection’ error. The fix changes the test to have l1 call connect() to l2 instead, avoiding the simultaneous-connect race. No daemon, protocol, or library code is changed.
Changed components
tests/test_splicing.pyInspect captured patch +1 / −1
diff --git a/tests/test_splicing.py b/tests/test_splicing.py
index e5a3565b..f563ba17 100644
--- a/tests/test_splicing.py
+++ b/tests/test_splicing.py
@@ -551,7 +551,7 @@ def test_route_by_old_scid(node_factory, bitcoind):
# Now restart l2, make sure it remembers the original!
l2.restart()
- l2.rpc.connect(l1.info['id'], 'localhost', l1.port)
+ l1.rpc.connect(l2.info['id'], 'localhost', l2.port)
l2.rpc.connect(l3.info['id'], 'localhost', l3.port)
wait_for(lambda: only_one(l1.rpc.listpeers()['peers'])['connected'] is True)
Why this scored 15/100
Community notes
Notes can correct, qualify, or add evidence to the AI analysis. Every note shown here has been validated by a human moderator.
The AI analysis stands alone for now. Submit a note if you can add evidence or important context.