A local call waiting for a slot is reached by a stop, a delete or a cancel and ends without running; what a call emits after its flow was stopped is dropped as its result is; a node run by hand whose flow is stopped or deleted tells its caller but publishes no node_error; cancelling a node also ends its calls on workers
Docs / docs (push) Successful in 33s
Playwright Tests / test-playwright (1, 2) (push) Successful in 3m45s
Playwright Tests / test-playwright (2, 2) (push) Successful in 2m47s
pre-commit / pre-commit (push) Successful in 3m44s
Test Backend / test-backend (3.12) (push) Failing after 5m55s
Test Backend / test-backend (3.14) (push) Failing after 5m31s
Compose Smoke Test / test-compose (push) Successful in 33s
Playwright Tests / merge-reports (push) Successful in 1m19s

This commit is contained in:
2026-11-11 12:27:33 +01:00
parent 7132d26f36
commit 8dcceecd95
11 changed files with 219 additions and 35 deletions
+6 -5
View File
@@ -96,11 +96,12 @@ Source code travels with each call and compiles on the worker, so remote-only
dependencies do not need to be installed on the engine. Calls to `emit`,
`save_artifact` and `load_artifact` work there too.
Cancelling a run stops its active node processes. Stopping or deleting a flow
does the same for the calls it has out, and drops an answer that still comes
back. A disconnected worker fails in-flight calls. While executing, the agent
sends heartbeats every ten seconds; 90 seconds without contact fails the call.
Heartbeats do not reset the node's own silence timeout.
Cancelling a run or a single node stops its active node processes. Stopping
or deleting a flow does the same for the calls it has out, and drops an answer
or emission that still comes back. A disconnected worker fails in-flight
calls. While executing, the agent sends heartbeats every ten seconds; 90
seconds without contact fails the call. Heartbeats do not reset the node's own
silence timeout.
Every call runs in a process group of its own. The agent kills the group when
the engine ends the call, when the connection drops and when the agent stops,
+4 -2
View File
@@ -168,8 +168,10 @@ Build issues prevent the affected flow from running. `node_unhealthy` and
the node recovers. A degraded node is a warning on Home rather than a fault.
You can **stop** a live flow to remove its subscriptions and schedules and end
the node calls it has running, or **pause** it to hold pending work while
values still arrive. What a stopped call still returns is dropped. **Step**
the node calls it has running or queued, or **pause** it to hold pending work
while values still arrive. What a stopped call still emits or returns is
dropped. A node you started by hand reports the cancel to you and is not
marked as failed. **Step**
releases one held item. Repeated background crashes can quarantine a flow;
publishing a change gives it another chance.