starred/orca
1
0
Fork 0
mirror of https://github.com/stablyai/orca.git synced 2026-09-22 00:02:31 +00:00
Code Issues Packages Projects Releases Wiki Activity

fix(e2e): fabricate runtime-host health through the snapshot that owns it (#20762)

Browse Source 
* fix(e2e): seed the runtime host with a snapshot the host cannot overwrite

Since #20003 the published snapshot owns runtime-host health, so a bare status
seed no longer survives: main's status owner publishes `checking` for this
unreachable synthetic host as soon as any runtime RPC touches it, the host
reads `connecting`, and the Add Project dialog falls back to Local — so the
host-scoped copy the test asserts never renders.

Seed a snapshot pinned at the top sequence instead. applyRuntimeHostStatusSnapshot
drops any later publication whose sequence is not higher, and setRuntimeEnvironmentStatus
no-ops a snapshot-less write once a snapshot exists, so the bare seed could not
have worked either way.

Ablated: passes with the pinned snapshot, fails without it.

* fix(e2e): fabricate paired-web host health through the snapshot that owns it

Companion to the onboarding seed fix. Since #20003 the published snapshot owns
runtime-host health, so writing `status: null` alone leaves the paired-web
client's verified/ready snapshot in place: addRuntimeHost reads
snapshot.transport 'ready' before it ever consults status, and the host still
renders Connected while the test waits for Disconnected.

Rewrite the snapshot coherently instead, pinned at the top sequence so the live
status owner cannot restore the host mid-assertion. The disconnected leg uses
transport 'unknown' rather than 'disconnected', because a dropped transport is
unverifiable and renders as Connecting; only a never-reached host renders
Disconnected. remoteControl is nulled because runtimeControlHealth answers
'available' on a ready control socket even with a null status.

Ablated with ORCA_E2E_WEB_CLIENT=1 (the whole file is test.skip'd without it,
so a run without that flag reports a passing skip): fails without the change,
passes with it.

* test(e2e): preserve paired runtime status metadata
This commit is contained in:
Brennan Benson
2026-09-15 16:18:36 -07:00
committed by GitHub
parent 0869d997c8
commit b5a99462bc
2 changed files with 58 additions and 20 deletions
Download Patch File Download Diff File Expand all files Collapse all files
tests/e2e/onboarding.spec.ts
+24 -9
View File
@@ -452,17 +452,32 @@ test.describe('Onboarding flow', () => {
// has a live, protocol-compatible status; without one it reads
// 'disconnected' and the Add Project dialog falls back to Local Mac.
// runtimeProtocolVersion 3 clears MIN_COMPATIBLE_RUNTIME_SERVER_VERSION.
const seededStatus = {
runtimeId: `${environment.id}-runtime`,
rendererGraphEpoch: 0,
graphStatus: 'ready' as const,
authoritativeWindowId: null,
liveTabCount: 0,
liveLeafCount: 0,
runtimeProtocolVersion: 3,
minCompatibleRuntimeClientVersion: 1
}
// Why a snapshot and not a bare status: since #20003 the published snapshot owns host health,
// and main's status owner publishes `checking` for this unreachable host as soon as any
// runtime RPC touches it — which flips the Add Project host to Local mid-test. Pinning the
// seed at the top sequence makes applyRuntimeHostStatusSnapshot's monotonic guard drop those
// publications. A snapshot-less write would also no-op once any snapshot exists.
store.getState().setRuntimeEnvironmentStatus(environment.id, {
status: {
runtimeId: `${environment.id}-runtime`,
rendererGraphEpoch: 0,
graphStatus: 'ready',
authoritativeWindowId: null,
liveTabCount: 0,
liveLeafCount: 0,
runtimeProtocolVersion: 3,
minCompatibleRuntimeClientVersion: 1
snapshot: {
environmentId: environment.id,
pairingRevision: environment.pairingRevision ?? environment.createdAt,
sequence: Number.MAX_SAFE_INTEGER,
checkedAt: Date.now(),
status: seededStatus,
verification: 'verified',
transport: 'ready'
},
status: seededStatus,
checkedAt: Date.now()
})
// Why: the store's switchRuntimeEnvironment probes reachability, which a
tests/e2e/paired-web-add-project-unavailable-host.spec.ts
+34 -11
View File
@@ -35,20 +35,43 @@ async function setOnlyRuntimeHostHealth(page: Page, health: HostHealth): Promise
if (nextHealth === 'blocked' && !current?.status) {
throw new Error('Paired web runtime status unavailable for compatibility fault')
}
// Why rewrite the snapshot and not just the status: since #20003 the published snapshot owns
// host health. A bare `status: null` leaves the paired-web client's verified/ready snapshot in
// place, addRuntimeHost reads transport 'ready' before it ever looks at status, and the host
// still renders Connected. Pinning at the top sequence also stops the live status owner
// restoring the host mid-assertion.
const verified = current?.status ?? current?.snapshot?.status ?? null
const snapshot = {
environmentId: environment.id,
pairingRevision: environment.pairingRevision ?? environment.createdAt,
sequence: Number.MAX_SAFE_INTEGER,
checkedAt: Date.now(),
...(nextHealth === 'blocked'
? {
status: { ...verified!, protocolVersion: 0, runtimeProtocolVersion: 0 },
verification: 'verified' as const,
transport: 'ready' as const
}
: {
status: null,
verification: 'unavailable' as const,
// Why 'unknown' and not 'disconnected': an unavailable/unknown snapshot falls
// through to disconnected health; explicit disconnected transport is treated as
// reconnecting and renders Connecting.
transport: 'unknown' as const
})
}
store.setState({
runtimeStatusByEnvironmentId: new Map(state.runtimeStatusByEnvironmentId).set(
environment.id,
nextHealth === 'blocked'
? {
...current,
checkedAt: Date.now(),
status: {
...current!.status!,
protocolVersion: 0,
runtimeProtocolVersion: 0
}
}
: { ...current, checkedAt: Date.now(), status: null }
{
...current,
snapshot,
checkedAt: snapshot.checkedAt,
status: snapshot.verification === 'verified' ? snapshot.status : null,
// Why: runtimeHealth answers 'available' on a ready remoteControl even with a null status.
remoteControl: null
}
)
})
return environment.name