rollbridge 0.1.35 → 0.1.37
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/README.md +5 -3
- package/changelog.d/20260830-guardian-retired-replacement-process-key.md +10 -3
- package/docs/cli.md +2 -1
- package/docs/config.md +4 -2
- package/docs/troubleshooting.md +3 -0
- package/package.json +1 -1
- package/src/daemon.js +226 -60
- package/src/guardian-client.js +99 -1
- package/src/process-guardian.js +31 -6
- package/test/fixtures/partial-owner-replacement-process-guardian.js +106 -0
- package/test/guardian-client.test.js +97 -0
- package/test/owner-replacement.test.js +604 -8
package/README.md
CHANGED
|
@@ -240,9 +240,11 @@ later deploys continue to honor the original drain boundary.
|
|
|
240
240
|
|
|
241
241
|
There is one explicit compatibility boundary: the first upgrade from a genuine
|
|
242
242
|
pre-owner-replacement Rollbridge guardian and daemon cannot share its listeners
|
|
243
|
-
on supported Node 20.
|
|
244
|
-
|
|
245
|
-
|
|
243
|
+
on supported Node 20. The same boundary applies to a retained transition guardian
|
|
244
|
+
that supports prepare/stage but predates the retired-owner commit command. After
|
|
245
|
+
an explicit capability probe and authentication of the guardian, exact daemon
|
|
246
|
+
PID/socket, runtime authority, and durable owned-process state, `ensure-daemon`
|
|
247
|
+
performs a one-time **disruptive** bridge. Existing proxy/control connections may close,
|
|
246
248
|
the retained processes keep their exact PIDs under guardian supervision, and
|
|
247
249
|
`status.ownerTransition` reports `mode: "legacy-first-upgrade"` with
|
|
248
250
|
`disruptive: true`. The bridge requires the existing config identity unchanged;
|
|
@@ -1,9 +1,16 @@
|
|
|
1
1
|
### Fixed
|
|
2
2
|
|
|
3
3
|
- Select retired-owner commit proof from the guardian-published committed owner
|
|
4
|
-
snapshot rather than candidate-local process order,
|
|
5
|
-
|
|
6
|
-
authority, and control-path fences.
|
|
4
|
+
snapshot rather than candidate-local process order, reserve its incumbent-owned
|
|
5
|
+
registration through commit, then attach it to the committed candidate while
|
|
6
|
+
retaining the replacement transaction, authority, and control-path fences.
|
|
7
7
|
- Fail closed when an older retained guardian cannot commit that replacement
|
|
8
8
|
atomically after the incumbent control socket disappears, preserving the
|
|
9
9
|
incumbent owner, retained connections, and guardian-managed release processes.
|
|
10
|
+
- Explicitly classify retained guardian replacement capabilities before preparing
|
|
11
|
+
a transaction. Guardians with prepare/stage support but no retired-owner commit
|
|
12
|
+
command use the fully attested one-time disruptive legacy upgrade bridge;
|
|
13
|
+
malformed, stale, or ambiguous protocol responses continue to fail closed. The
|
|
14
|
+
partial guardian's prepared transaction remains the mutation fence through
|
|
15
|
+
candidate reconstruction and boundary revalidation, and failed preparation
|
|
16
|
+
notifies the incumbent to resume paused release drains.
|
package/docs/cli.md
CHANGED
|
@@ -103,7 +103,8 @@ the candidate binds and validates its listeners, and an authenticated fenced
|
|
|
103
103
|
transaction commits guardian authority and the final control socket. A lost
|
|
104
104
|
control response is accepted only when the guardian confirms the exact committed
|
|
105
105
|
transaction id. The first authenticated upgrade from a genuine pre-replacement
|
|
106
|
-
guardian/daemon
|
|
106
|
+
guardian/daemon, or from a retained prepare/stage guardian that explicitly lacks
|
|
107
|
+
the retired-owner commit capability, is the sole exception: `ensure-daemon` preserves its exact
|
|
107
108
|
guardian-owned processes but deliberately retires the old listeners before the
|
|
108
109
|
Node 20 candidate binds, so existing proxy/control connections may close.
|
|
109
110
|
Successful status JSON records this as `ownerTransition.disruptive: true` and
|
package/docs/config.md
CHANGED
|
@@ -151,8 +151,10 @@ config reloads remain unchanged.
|
|
|
151
151
|
|
|
152
152
|
The first upgrade from an authenticated pre-replacement Rollbridge guardian and
|
|
153
153
|
daemon uses an explicitly disruptive compatibility bridge because that legacy
|
|
154
|
-
owner cannot transfer listeners on Node 20.
|
|
155
|
-
|
|
154
|
+
owner cannot transfer listeners on Node 20. A retained guardian that supports
|
|
155
|
+
replacement prepare/stage but lacks the retired-owner commit command is explicitly
|
|
156
|
+
classified and uses the same bridge. Rollbridge attests the exact guardian and
|
|
157
|
+
daemon processes, sockets, runtime/config authority, and durable process
|
|
156
158
|
registrations before retiring the legacy listeners. Managed process PIDs and
|
|
157
159
|
release state remain supervised, but live proxy/control connections may close.
|
|
158
160
|
The resulting status includes `ownerTransition: {disruptive: true, mode:
|
package/docs/troubleshooting.md
CHANGED
|
@@ -11,6 +11,9 @@ cannot authenticate and attest an allowed owner transition.
|
|
|
11
11
|
|
|
12
12
|
**Fix.** With `ownerRecovery`, a genuine pre-owner-replacement Rollbridge daemon
|
|
13
13
|
and guardian can cross the documented one-time disruptive bridge automatically.
|
|
14
|
+
A retained guardian that implements replacement prepare/stage but lacks the
|
|
15
|
+
retired-owner commit capability uses that same bridge only after the exact
|
|
16
|
+
non-mutating capability signature and legacy process dispatch are confirmed.
|
|
14
17
|
Its existing proxy/control connections may close; successful status reports
|
|
15
18
|
`ownerTransition.mode: "legacy-first-upgrade"`. Keep the incumbent config
|
|
16
19
|
identity unchanged for that first invocation, then apply config/socket changes
|
package/package.json
CHANGED
package/src/daemon.js
CHANGED
|
@@ -29,7 +29,7 @@ const STATE_PERSIST_INTERVAL_MS = 5000
|
|
|
29
29
|
* @typedef {{configDigest: string, format: number, guardian: {pid?: number, socketPath: string, token: string}, reconnectGraceMs: number}} OwnerRecoveryMetadata
|
|
30
30
|
* @typedef {DaemonStatus & {recovery: OwnerRecoveryMetadata, singletonReleaseIds?: Record<string, string>}} OwnerRecoverySnapshot
|
|
31
31
|
* @typedef {{authority: JsonValue, config: import("./config.js").RollbridgeConfig, releaseConfigs?: Record<string, import("./config.js").RollbridgeConfig>, singletonReleaseIds?: Record<string, string>, snapshot: OwnerRecoverySnapshot}} PrivateOwnerState
|
|
32
|
-
* @typedef {{boundaryCrossed: boolean, incumbentControl
|
|
32
|
+
* @typedef {{boundaryCrossed: boolean, incumbentControl?: Awaited<ReturnType<typeof openControlSession>>, incumbentStartTime: string, legacyGuardian: GuardianClient, legacyInventory: {key: string, provenance: string}[], legacyPrepared?: {ownerState: JsonValue, replacementId: string}, legacySnapshot: OwnerRecoverySnapshot, prepared: {ownerState: JsonValue, replacementId: string}, recoverySnapshot: OwnerRecoverySnapshot}} LegacyOwnerBridge
|
|
33
33
|
*/
|
|
34
34
|
|
|
35
35
|
export default class RollbridgeDaemon {
|
|
@@ -325,24 +325,38 @@ export default class RollbridgeDaemon {
|
|
|
325
325
|
configDigest: persisted.recovery.configDigest,
|
|
326
326
|
runtime: persisted.daemonRuntime ? {...persisted.daemonRuntime} : null
|
|
327
327
|
}
|
|
328
|
-
|
|
329
|
-
|
|
328
|
+
const replacementProtocol = await this.guardian.ownerReplacementProtocol()
|
|
329
|
+
const legacyBridge = replacementProtocol === "legacy"
|
|
330
|
+
? await this.prepareLegacyOwnerReplacement({persisted, persistedAuthority})
|
|
331
|
+
: undefined
|
|
332
|
+
const prepared = legacyBridge?.prepared ?? await this.guardian.prepareOwnerReplacement(persistedAuthority, this.ownerAuthority())
|
|
333
|
+
let preparedStatus
|
|
334
|
+
let reservedProcessKey
|
|
335
|
+
let transfer
|
|
330
336
|
|
|
331
337
|
try {
|
|
332
|
-
|
|
338
|
+
preparedStatus = await this.guardian.replacementStatus()
|
|
339
|
+
transfer = /** @type {PrivateOwnerState} */ (prepared.ownerState)
|
|
340
|
+
if (!transfer?.config || !transfer.snapshot) throw new Error("Committed owner published incomplete replacement state")
|
|
341
|
+
const registeredProcesses = new Map((await this.guardian.inventory()).map(({key, provenance}) => [key, provenance]))
|
|
342
|
+
|
|
343
|
+
reservedProcessKey = legacyBridge ? undefined : ownerSnapshotProcessKeys(transfer.snapshot, transfer.singletonReleaseIds)
|
|
344
|
+
.find((key) => registeredProcesses.has(key))
|
|
345
|
+
if (reservedProcessKey) this.guardian.reserveProcessRecovery(reservedProcessKey, /** @type {string} */ (registeredProcesses.get(reservedProcessKey)))
|
|
346
|
+
await this.restoreOwnerState(transfer.snapshot, {config: transfer.config, releaseConfigs: transfer.releaseConfigs, resumeDrains: false, singletonReleaseIds: transfer.singletonReleaseIds, synchronizeLifecycleRoles: false})
|
|
333
347
|
} catch (error) {
|
|
334
|
-
legacyBridge
|
|
335
|
-
|
|
336
|
-
|
|
337
|
-
|
|
338
|
-
|
|
339
|
-
|
|
348
|
+
legacyBridge?.incumbentControl?.close()
|
|
349
|
+
if (legacyBridge) {
|
|
350
|
+
try {
|
|
351
|
+
await this.abandonLegacyOwnerBridge(legacyBridge)
|
|
352
|
+
} catch (cleanupError) {
|
|
353
|
+
throw new AggregateError([error, cleanupError], "Legacy owner replacement reconstruction failed and its upgrade coordinator could not be abandoned", {cause: cleanupError})
|
|
354
|
+
}
|
|
355
|
+
} else {
|
|
356
|
+
this.guardian.disconnect()
|
|
357
|
+
}
|
|
358
|
+
throw error
|
|
340
359
|
}
|
|
341
|
-
const preparedStatus = await this.guardian.replacementStatus()
|
|
342
|
-
const transfer = /** @type {PrivateOwnerState} */ (prepared.ownerState)
|
|
343
|
-
|
|
344
|
-
if (!transfer?.config || !transfer.snapshot) throw new Error("Committed owner published incomplete replacement state")
|
|
345
|
-
await this.restoreOwnerState(transfer.snapshot, {config: transfer.config, releaseConfigs: transfer.releaseConfigs, resumeDrains: false, singletonReleaseIds: transfer.singletonReleaseIds, synchronizeLifecycleRoles: false})
|
|
346
360
|
for (const release of this.releases.values()) release.preserveConfigOnRetirement = true
|
|
347
361
|
this.logger("owner replacement candidate prepared", {activeReleaseId: this.activeRelease?.releaseId ?? null, replacementId: prepared.replacementId})
|
|
348
362
|
|
|
@@ -386,6 +400,7 @@ export default class RollbridgeDaemon {
|
|
|
386
400
|
if (incumbentControl) {
|
|
387
401
|
const listenerSession = incumbentControl
|
|
388
402
|
|
|
403
|
+
if (reservedProcessKey) await this.guardian.recoverReservedProcess(reservedProcessKey)
|
|
389
404
|
incumbentControl.onEvent((event) => this.handleIncumbentListenerEvent(event, listenerSession))
|
|
390
405
|
await incumbentControl.request({
|
|
391
406
|
command: "yield-owner-listeners",
|
|
@@ -414,15 +429,20 @@ export default class RollbridgeDaemon {
|
|
|
414
429
|
snapshot: this.status()
|
|
415
430
|
})
|
|
416
431
|
|
|
432
|
+
if (staged.committed && reservedProcessKey) {
|
|
433
|
+
committedAuthority = true
|
|
434
|
+
await this.guardian.recoverReservedProcess(reservedProcessKey)
|
|
435
|
+
}
|
|
436
|
+
|
|
417
437
|
if (!staged.committed) {
|
|
418
438
|
if (retiredIncumbentControl) {
|
|
419
|
-
const
|
|
420
|
-
const processKey = ownerSnapshotProcessKeys(transfer.snapshot, transfer.singletonReleaseIds)
|
|
421
|
-
.find((key) => recoveredProcesses.has(key))
|
|
439
|
+
const processKey = reservedProcessKey
|
|
422
440
|
|
|
423
|
-
if (!processKey) throw new Error("Retired owner replacement requires an exact
|
|
441
|
+
if (!processKey || !this.guardian.processes.has(processKey)) throw new Error("Retired owner replacement requires an exact reserved process from committed owner state")
|
|
424
442
|
try {
|
|
425
443
|
await this.guardian.commitRetiredOwnerReplacement(prepared.replacementId, processKey)
|
|
444
|
+
committedAuthority = true
|
|
445
|
+
await this.guardian.recoverReservedProcess(processKey)
|
|
426
446
|
} catch (error) {
|
|
427
447
|
if (!(error instanceof Error) || error.message !== "Guardian commit-retired-owner-replacement requires the committed owner") throw error
|
|
428
448
|
throw new Error(
|
|
@@ -484,12 +504,13 @@ export default class RollbridgeDaemon {
|
|
|
484
504
|
}
|
|
485
505
|
if (legacyBridge && !legacyBridge.boundaryCrossed) {
|
|
486
506
|
try {
|
|
487
|
-
await this.
|
|
507
|
+
await this.abandonLegacyOwnerBridge(legacyBridge)
|
|
488
508
|
} catch (failure) {
|
|
489
509
|
abortError = failure instanceof Error ? failure : new Error(String(failure))
|
|
490
510
|
}
|
|
491
511
|
} else {
|
|
492
512
|
this.guardian.disconnect()
|
|
513
|
+
legacyBridge?.legacyGuardian.disconnect()
|
|
493
514
|
}
|
|
494
515
|
if (committed) {
|
|
495
516
|
const commitmentError = await committed
|
|
@@ -540,56 +561,74 @@ export default class RollbridgeDaemon {
|
|
|
540
561
|
}
|
|
541
562
|
|
|
542
563
|
/**
|
|
543
|
-
* Authenticates and prepares the one-time disruptive bridge for
|
|
544
|
-
* @param {{
|
|
564
|
+
* Authenticates and prepares the one-time disruptive bridge for an exact legacy guardian protocol.
|
|
565
|
+
* @param {{persisted: OwnerRecoverySnapshot, persistedAuthority: {configDigest: string, runtime: import("./daemon-runtime.js").DaemonRuntimeIdentity | null}}} options - Legacy evidence.
|
|
545
566
|
* @returns {Promise<LegacyOwnerBridge>} Prepared bridge.
|
|
546
567
|
*/
|
|
547
|
-
async prepareLegacyOwnerReplacement({
|
|
548
|
-
const diagnostic = error instanceof Error ? error.message : String(error)
|
|
568
|
+
async prepareLegacyOwnerReplacement({persisted, persistedAuthority}) {
|
|
549
569
|
const legacyProcessKey = ownerSnapshotProcessKeys(persisted, persisted.singletonReleaseIds)[0]
|
|
570
|
+
const legacyGuardian = this.guardian
|
|
571
|
+
let legacyPrepared
|
|
572
|
+
|
|
573
|
+
if (!legacyProcessKey) throw new Error("Legacy disruptive owner replacement requires an exact guardian-owned process registration in durable state")
|
|
574
|
+
if (!legacyGuardian) throw new Error("Legacy disruptive replacement is missing its authenticated guardian connection")
|
|
575
|
+
try {
|
|
576
|
+
legacyPrepared = await legacyGuardian.prepareOwnerReplacement(persistedAuthority, this.ownerAuthority())
|
|
577
|
+
} catch (error) {
|
|
578
|
+
const diagnostic = error instanceof Error ? error.message : String(error)
|
|
579
|
+
|
|
580
|
+
if (!isLegacyGuardianPrepareDiagnostic(diagnostic)) throw error
|
|
581
|
+
let probeAccepted = false
|
|
550
582
|
|
|
551
|
-
if (!isLegacyGuardianPrepareDiagnostic(diagnostic)) throw error
|
|
552
|
-
if (!legacyProcessKey) throw new Error("Legacy disruptive owner replacement requires an exact guardian-owned process registration in durable state", {cause: error})
|
|
553
|
-
if (diagnostic !== "Unknown guardian command: prepare-owner-replacement") {
|
|
554
583
|
try {
|
|
555
|
-
await
|
|
584
|
+
await legacyGuardian.request({
|
|
556
585
|
authority: persistedAuthority,
|
|
557
586
|
command: "prepare-owner-replacement",
|
|
558
587
|
key: legacyProcessKey,
|
|
559
588
|
nextAuthority: this.ownerAuthority()
|
|
560
589
|
})
|
|
561
|
-
|
|
590
|
+
probeAccepted = true
|
|
562
591
|
} catch (probeError) {
|
|
563
592
|
if (!(probeError instanceof Error) || probeError.message !== "Unknown guardian command: prepare-owner-replacement") {
|
|
564
593
|
throw new Error("Guardian does not match the authenticated pre-split replacement protocol signature", {cause: probeError})
|
|
565
594
|
}
|
|
566
595
|
}
|
|
596
|
+
if (probeAccepted) throw new Error("Legacy guardian protocol probe unexpectedly accepted owner replacement", {cause: error})
|
|
567
597
|
}
|
|
568
|
-
|
|
569
|
-
throw new Error("The one-time legacy guardian bridge requires the incumbent config identity unchanged; retry the config change after the bridge establishes split-3 authority")
|
|
570
|
-
}
|
|
571
|
-
if (!this.legacyIncumbentPid) throw new Error("The authenticated pre-split guardian requires an exact incumbent PID from ensure-daemon for the disruptive bridge")
|
|
572
|
-
if (!this.guardianIdentity?.pid) throw new Error("The authenticated pre-split guardian state is missing its exact guardian PID")
|
|
573
|
-
const legacyGuardian = this.guardian
|
|
574
|
-
|
|
575
|
-
if (!legacyGuardian) throw new Error("Legacy disruptive replacement is missing its authenticated guardian connection")
|
|
576
|
-
|
|
577
|
-
await verifyLegacyGuardianProcess(this.guardianIdentity.pid, this.guardianIdentity.socketPath)
|
|
578
|
-
const incumbentStartTime = await verifyLegacyDaemonProcess(this.legacyIncumbentPid, this.configPath, persisted.control.path)
|
|
579
|
-
const incumbentControl = await openControlSession(persisted.control.path)
|
|
598
|
+
let incumbentControl
|
|
580
599
|
let upgraded
|
|
581
600
|
|
|
582
601
|
try {
|
|
583
|
-
|
|
602
|
+
if (persisted.recovery.configDigest !== this.ownerRecoveryConfigDigest()) {
|
|
603
|
+
throw new Error("The one-time legacy guardian bridge requires the incumbent config identity unchanged; retry the config change after the bridge establishes split-3 authority")
|
|
604
|
+
}
|
|
605
|
+
if (!this.legacyIncumbentPid) throw new Error("The authenticated legacy guardian requires an exact incumbent PID from ensure-daemon for the disruptive bridge")
|
|
606
|
+
if (!this.guardianIdentity?.pid) throw new Error("The authenticated legacy guardian state is missing its exact guardian PID")
|
|
607
|
+
await verifyLegacyGuardianProcess(this.guardianIdentity.pid, this.guardianIdentity.socketPath)
|
|
608
|
+
const incumbentStartTime = await verifyLegacyDaemonProcess(this.legacyIncumbentPid, this.configPath, persisted.control.path)
|
|
609
|
+
if (legacyPrepared) assertLegacyPrivateOwnerState(legacyPrepared.ownerState, persisted, persistedAuthority)
|
|
610
|
+
try {
|
|
611
|
+
incumbentControl = await openControlSession(persisted.control.path)
|
|
612
|
+
} catch (error) {
|
|
613
|
+
if (!legacyPrepared || !error || typeof error !== "object" || !("code" in error) || error.code !== "ENOENT") throw error
|
|
614
|
+
}
|
|
615
|
+
if (incumbentControl) {
|
|
616
|
+
const incumbentStatus = await incumbentControl.request({command: "status"})
|
|
584
617
|
|
|
585
|
-
|
|
586
|
-
const ownerState = {
|
|
587
|
-
authority: persistedAuthority,
|
|
588
|
-
config: this.config,
|
|
589
|
-
releaseConfigs: Object.fromEntries(persisted.releases.map((release) => [release.releaseId, this.config])),
|
|
590
|
-
singletonReleaseIds: Object.fromEntries(persisted.singletons.map((singleton) => [singleton.id, persisted.activeReleaseId]).filter((entry) => entry[1] !== null)),
|
|
591
|
-
snapshot: persisted
|
|
618
|
+
assertLegacyIncumbentStatus(incumbentStatus, persisted)
|
|
592
619
|
}
|
|
620
|
+
const ownerState = legacyPrepared
|
|
621
|
+
? legacyPrepared.ownerState
|
|
622
|
+
: {
|
|
623
|
+
authority: persistedAuthority,
|
|
624
|
+
config: this.config,
|
|
625
|
+
releaseConfigs: Object.fromEntries(persisted.releases.map((release) => [release.releaseId, this.config])),
|
|
626
|
+
singletonReleaseIds: Object.fromEntries(persisted.singletons.map((singleton) => [singleton.id, persisted.activeReleaseId]).filter((entry) => entry[1] !== null)),
|
|
627
|
+
snapshot: persisted
|
|
628
|
+
}
|
|
629
|
+
const legacyInventory = normalizeLegacyGuardianInventory(await legacyGuardian.inventory())
|
|
630
|
+
|
|
631
|
+
assertLegacyGuardianInventoryMembership(legacyInventory, /** @type {PrivateOwnerState} */ (ownerState))
|
|
593
632
|
const upgradedIdentity = /** @type {{pid?: number, socketPath: string, token: string}} */ ({
|
|
594
633
|
socketPath: `${this.statePath}.split3-guardian.sock`,
|
|
595
634
|
token: crypto.randomBytes(32).toString("hex")
|
|
@@ -598,7 +637,6 @@ export default class RollbridgeDaemon {
|
|
|
598
637
|
await assertPathAbsent(upgradedIdentity.socketPath, "Legacy upgrade guardian socket")
|
|
599
638
|
upgraded = await legacyGuardian.upgradeLegacyGuardian({ownerState, ...upgradedIdentity})
|
|
600
639
|
upgradedIdentity.pid = upgraded.pid
|
|
601
|
-
legacyGuardian.disconnect()
|
|
602
640
|
this.guardian = upgraded
|
|
603
641
|
this.guardianIdentity = upgradedIdentity
|
|
604
642
|
const prepared = await upgraded.prepareOwnerReplacement(persistedAuthority, this.ownerAuthority())
|
|
@@ -612,36 +650,102 @@ export default class RollbridgeDaemon {
|
|
|
612
650
|
incumbentPid: this.legacyIncumbentPid,
|
|
613
651
|
replacementId: prepared.replacementId
|
|
614
652
|
})
|
|
615
|
-
return {boundaryCrossed: false, incumbentControl, incumbentStartTime, prepared, recoverySnapshot}
|
|
653
|
+
return {boundaryCrossed: false, incumbentControl, incumbentStartTime, legacyGuardian, legacyInventory, legacyPrepared, legacySnapshot: persisted, prepared, recoverySnapshot}
|
|
616
654
|
} catch (upgradeError) {
|
|
617
|
-
incumbentControl
|
|
618
|
-
|
|
655
|
+
incumbentControl?.close()
|
|
656
|
+
let cleanupError
|
|
657
|
+
|
|
658
|
+
if (upgraded) {
|
|
659
|
+
try {
|
|
660
|
+
await upgraded.abandonLegacyUpgrade()
|
|
661
|
+
} catch (error) {
|
|
662
|
+
cleanupError = error instanceof Error ? error : new Error(String(error))
|
|
663
|
+
}
|
|
664
|
+
}
|
|
665
|
+
legacyGuardian.disconnect()
|
|
666
|
+
if (cleanupError) {
|
|
667
|
+
throw new AggregateError([upgradeError, cleanupError], "Legacy owner replacement validation failed and its upgrade coordinator could not be abandoned", {cause: upgradeError})
|
|
668
|
+
}
|
|
619
669
|
throw upgradeError
|
|
620
670
|
}
|
|
621
671
|
}
|
|
622
672
|
|
|
673
|
+
/**
|
|
674
|
+
* Abandons the uncommitted coordinator before releasing the incumbent guardian transaction.
|
|
675
|
+
* @param {LegacyOwnerBridge} bridge - Prepared bridge.
|
|
676
|
+
*/
|
|
677
|
+
async abandonLegacyOwnerBridge(bridge) {
|
|
678
|
+
let cleanupError
|
|
679
|
+
|
|
680
|
+
try {
|
|
681
|
+
await this.guardian?.abandonLegacyUpgrade()
|
|
682
|
+
} catch (error) {
|
|
683
|
+
cleanupError = error instanceof Error ? error : new Error(String(error))
|
|
684
|
+
}
|
|
685
|
+
bridge.legacyGuardian.disconnect()
|
|
686
|
+
if (cleanupError) throw cleanupError
|
|
687
|
+
}
|
|
688
|
+
|
|
623
689
|
/**
|
|
624
690
|
* Crosses the explicitly disruptive legacy-only boundary after candidate reconstruction.
|
|
625
691
|
* @param {LegacyOwnerBridge} bridge - Prepared bridge.
|
|
626
692
|
*/
|
|
627
693
|
async crossLegacyDisruptiveBoundary(bridge) {
|
|
628
694
|
if (!this.legacyIncumbentPid || !this.statePath) throw new Error("Legacy disruptive boundary is missing its exact incumbent identity")
|
|
695
|
+
if (!this.guardian) throw new Error("Legacy disruptive boundary is missing its authenticated upgraded guardian")
|
|
696
|
+
if (!bridge.legacyGuardian.pid) throw new Error("Legacy disruptive boundary is missing its authenticated retained guardian PID")
|
|
697
|
+
await verifyLegacyGuardianProcess(bridge.legacyGuardian.pid, bridge.legacyGuardian.socketPath)
|
|
629
698
|
const currentStartTime = await verifyLegacyDaemonProcess(this.legacyIncumbentPid, this.configPath, bridge.recoverySnapshot.control.path)
|
|
630
699
|
|
|
631
700
|
if (currentStartTime !== bridge.incumbentStartTime) throw new Error("Legacy incumbent PID identity changed before the disruptive boundary")
|
|
701
|
+
if (!isDeepStrictEqual(await readState(this.statePath), bridge.legacySnapshot)) {
|
|
702
|
+
throw new Error("Legacy owner recovery state changed before the disruptive boundary")
|
|
703
|
+
}
|
|
704
|
+
if (bridge.incumbentControl) {
|
|
705
|
+
assertLegacyIncumbentStatus(await bridge.incumbentControl.request({command: "status"}), bridge.legacySnapshot)
|
|
706
|
+
}
|
|
707
|
+
const currentInventory = normalizeLegacyGuardianInventory(await bridge.legacyGuardian.inventory())
|
|
708
|
+
|
|
709
|
+
if (!isDeepStrictEqual(currentInventory, bridge.legacyInventory)) {
|
|
710
|
+
throw new Error("Legacy guardian process inventory changed before the disruptive boundary")
|
|
711
|
+
}
|
|
712
|
+
const candidateOwnerState = {
|
|
713
|
+
authority: this.ownerAuthority(),
|
|
714
|
+
config: this.config,
|
|
715
|
+
releaseConfigs: Object.fromEntries([...this.releases].map(([releaseId, release]) => [releaseId, release.config])),
|
|
716
|
+
singletonReleaseIds: Object.fromEntries(this.singletonReleaseIds),
|
|
717
|
+
snapshot: this.status()
|
|
718
|
+
}
|
|
719
|
+
|
|
720
|
+
assertLegacyGuardianInventoryMembership(currentInventory, candidateOwnerState)
|
|
721
|
+
let legacyCommitted
|
|
722
|
+
|
|
723
|
+
if (bridge.legacyPrepared) {
|
|
724
|
+
const status = await bridge.legacyGuardian.replacementStatus()
|
|
725
|
+
|
|
726
|
+
if (!status.ownerClaimed) throw new Error("Legacy guardian incumbent ownership changed before the disruptive boundary")
|
|
727
|
+
legacyCommitted = bridge.legacyGuardian.waitForEvent("replacement-committed")
|
|
728
|
+
const staged = await bridge.legacyGuardian.stageOwnerReplacement(bridge.legacyPrepared.replacementId, candidateOwnerState)
|
|
729
|
+
|
|
730
|
+
if (staged.committed) throw new Error("Legacy guardian committed replacement before the authenticated disruptive boundary")
|
|
731
|
+
}
|
|
632
732
|
this.logger("legacy owner replacement disruptive boundary", {
|
|
633
733
|
disruptive: true,
|
|
634
734
|
incumbentPid: this.legacyIncumbentPid,
|
|
635
|
-
reason: "
|
|
735
|
+
reason: "retained guardian and daemon lacked atomic replacement protocol"
|
|
636
736
|
})
|
|
737
|
+
await this.guardian.beginLegacyOwnerClaim(bridge.prepared.replacementId, bridge.recoverySnapshot.recovery.reconnectGraceMs)
|
|
637
738
|
process.kill(this.legacyIncumbentPid, "SIGKILL")
|
|
638
739
|
bridge.boundaryCrossed = true
|
|
639
|
-
await bridge.incumbentControl
|
|
740
|
+
await bridge.incumbentControl?.closed()
|
|
741
|
+
if (legacyCommitted) await legacyCommitted
|
|
742
|
+
bridge.legacyGuardian.disconnect()
|
|
743
|
+
await this.guardian.completeLegacyOwnerClaim(bridge.prepared.replacementId)
|
|
640
744
|
await writeState(this.statePath, bridge.recoverySnapshot)
|
|
641
745
|
this.ownerTransition = /** @type {OwnerTransition} */ ({
|
|
642
746
|
disruptive: true,
|
|
643
747
|
mode: "legacy-first-upgrade",
|
|
644
|
-
reason: "
|
|
748
|
+
reason: "retained guardian and daemon lacked atomic replacement protocol"
|
|
645
749
|
})
|
|
646
750
|
}
|
|
647
751
|
|
|
@@ -1998,7 +2102,7 @@ export function isLegacyGuardianPrepareDiagnostic(diagnostic) {
|
|
|
1998
2102
|
}
|
|
1999
2103
|
|
|
2000
2104
|
/**
|
|
2001
|
-
* @param {
|
|
2105
|
+
* @param {DaemonStatus} snapshot - Serialized owner process snapshot.
|
|
2002
2106
|
* @param {Record<string, string>} [singletonReleaseIds] - Exact singleton owner releases.
|
|
2003
2107
|
* @returns {string[]} Exact guardian registration keys present in the snapshot.
|
|
2004
2108
|
*/
|
|
@@ -2017,6 +2121,40 @@ function ownerSnapshotProcessKeys(snapshot, singletonReleaseIds = {}) {
|
|
|
2017
2121
|
return keys
|
|
2018
2122
|
}
|
|
2019
2123
|
|
|
2124
|
+
/**
|
|
2125
|
+
* Canonicalizes the authenticated guardian inventory without relying on map insertion order.
|
|
2126
|
+
* @param {{key: string, provenance: string}[]} inventory - Guardian inventory response.
|
|
2127
|
+
* @returns {{key: string, provenance: string}[]} Exact key/provenance fence.
|
|
2128
|
+
*/
|
|
2129
|
+
function normalizeLegacyGuardianInventory(inventory) {
|
|
2130
|
+
const normalized = inventory.map((entry) => {
|
|
2131
|
+
if (!entry || typeof entry.key !== "string" || !entry.key || typeof entry.provenance !== "string" || !entry.provenance) {
|
|
2132
|
+
throw new Error("Legacy guardian returned an invalid process inventory")
|
|
2133
|
+
}
|
|
2134
|
+
return {key: entry.key, provenance: entry.provenance}
|
|
2135
|
+
}).sort((left, right) => left.key.localeCompare(right.key))
|
|
2136
|
+
|
|
2137
|
+
if (new Set(normalized.map(({key}) => key)).size !== normalized.length) {
|
|
2138
|
+
throw new Error("Legacy guardian returned duplicate process registrations")
|
|
2139
|
+
}
|
|
2140
|
+
return normalized
|
|
2141
|
+
}
|
|
2142
|
+
|
|
2143
|
+
/**
|
|
2144
|
+
* Requires every and only the process registrations serialized by committed private owner state.
|
|
2145
|
+
* @param {{key: string, provenance: string}[]} inventory - Canonical guardian inventory.
|
|
2146
|
+
* @param {{snapshot: DaemonStatus, singletonReleaseIds?: Record<string, string>}} ownerState - Authenticated owner process snapshot.
|
|
2147
|
+
*/
|
|
2148
|
+
function assertLegacyGuardianInventoryMembership(inventory, ownerState) {
|
|
2149
|
+
if (!ownerState?.snapshot) throw new Error("Legacy guardian owner state is missing its committed process snapshot")
|
|
2150
|
+
const inventoryKeys = inventory.map(({key}) => key)
|
|
2151
|
+
const snapshotKeys = ownerSnapshotProcessKeys(ownerState.snapshot, ownerState.singletonReleaseIds).sort((left, right) => left.localeCompare(right))
|
|
2152
|
+
|
|
2153
|
+
if (!isDeepStrictEqual(inventoryKeys, snapshotKeys)) {
|
|
2154
|
+
throw new Error("Legacy guardian process inventory does not match committed owner state")
|
|
2155
|
+
}
|
|
2156
|
+
}
|
|
2157
|
+
|
|
2020
2158
|
/**
|
|
2021
2159
|
* Verifies the exact authenticated guardian process and socket without scanning other PIDs.
|
|
2022
2160
|
* @param {number} pid - Persisted guardian PID.
|
|
@@ -2026,7 +2164,7 @@ async function verifyLegacyGuardianProcess(pid, socketPath) {
|
|
|
2026
2164
|
const args = await processArguments(pid, "legacy guardian")
|
|
2027
2165
|
const script = args.find((argument) => argument.endsWith("process-guardian.js"))
|
|
2028
2166
|
|
|
2029
|
-
if (!script || !args.includes(socketPath)) throw new Error(`Persisted guardian PID ${pid} does not match the
|
|
2167
|
+
if (!script || !args.includes(socketPath)) throw new Error(`Persisted guardian PID ${pid} does not match the retained guardian command and socket`)
|
|
2030
2168
|
await verifyProcessUser(pid, "legacy guardian")
|
|
2031
2169
|
await verifyUnixSocketOwner(pid, socketPath, "legacy guardian")
|
|
2032
2170
|
}
|
|
@@ -2045,7 +2183,7 @@ async function verifyLegacyDaemonProcess(pid, configPath, socketPath) {
|
|
|
2045
2183
|
const configIndex = args.indexOf("--config")
|
|
2046
2184
|
|
|
2047
2185
|
if (daemonIndex < 0 || configIndex < 0 || args[configIndex + 1] !== configPath) {
|
|
2048
|
-
throw new Error(`Daemon PID ${pid} does not match the exact
|
|
2186
|
+
throw new Error(`Daemon PID ${pid} does not match the exact retained daemon config command`)
|
|
2049
2187
|
}
|
|
2050
2188
|
await verifyProcessUser(pid, "legacy daemon")
|
|
2051
2189
|
await verifyUnixSocketOwner(pid, socketPath, "legacy daemon")
|
|
@@ -2121,10 +2259,38 @@ function assertLegacyIncumbentStatus(status, persisted) {
|
|
|
2121
2259
|
if (status.application !== persisted.application || !control || typeof control !== "object" || Array.isArray(control) || control.path !== persisted.control.path ||
|
|
2122
2260
|
!runtime || typeof runtime !== "object" || Array.isArray(runtime) || runtime.digest !== persisted.daemonRuntime?.digest ||
|
|
2123
2261
|
!ownerRecovery || typeof ownerRecovery !== "object" || Array.isArray(ownerRecovery) || ownerRecovery.configDigest !== persisted.recovery.configDigest) {
|
|
2124
|
-
throw new Error("Responsive incumbent does not match the exact persisted
|
|
2262
|
+
throw new Error("Responsive incumbent does not match the exact persisted retained-daemon authority")
|
|
2125
2263
|
}
|
|
2126
2264
|
}
|
|
2127
2265
|
|
|
2266
|
+
/**
|
|
2267
|
+
* Validates private committed state returned by a partial transaction guardian.
|
|
2268
|
+
* @param {JsonValue} value - Authenticated guardian owner state.
|
|
2269
|
+
* @param {OwnerRecoverySnapshot} persisted - Durable expected identity.
|
|
2270
|
+
* @param {{configDigest: string, runtime: import("./daemon-runtime.js").DaemonRuntimeIdentity | null}} persistedAuthority - Exact committed authority.
|
|
2271
|
+
*/
|
|
2272
|
+
function assertLegacyPrivateOwnerState(value, persisted, persistedAuthority) {
|
|
2273
|
+
if (!value || typeof value !== "object" || Array.isArray(value)) throw new Error("Partial legacy guardian returned invalid committed owner state")
|
|
2274
|
+
const state = /** @type {Record<string, JsonValue>} */ (value)
|
|
2275
|
+
const config = state.config
|
|
2276
|
+
const snapshot = state.snapshot
|
|
2277
|
+
|
|
2278
|
+
if (!isDeepStrictEqual(state.authority, persistedAuthority) || !config || typeof config !== "object" || Array.isArray(config) ||
|
|
2279
|
+
ownerConfigDigest(/** @type {import("./config.js").RollbridgeConfig} */ (config)) !== persistedAuthority.configDigest ||
|
|
2280
|
+
!snapshot || typeof snapshot !== "object" || Array.isArray(snapshot)) {
|
|
2281
|
+
throw new Error("Partial legacy guardian committed owner state does not match the persisted authority")
|
|
2282
|
+
}
|
|
2283
|
+
const privateSnapshot = /** @type {OwnerRecoverySnapshot} */ (snapshot)
|
|
2284
|
+
|
|
2285
|
+
assertLegacyIncumbentStatus(/** @type {Record<string, JsonValue>} */ (snapshot), persisted)
|
|
2286
|
+
if (!isDeepStrictEqual(
|
|
2287
|
+
ownerSnapshotProcessKeys(privateSnapshot, state.singletonReleaseIds && typeof state.singletonReleaseIds === "object" && !Array.isArray(state.singletonReleaseIds)
|
|
2288
|
+
? /** @type {Record<string, string>} */ (state.singletonReleaseIds)
|
|
2289
|
+
: {}).sort(),
|
|
2290
|
+
ownerSnapshotProcessKeys(persisted, persisted.singletonReleaseIds).sort()
|
|
2291
|
+
)) throw new Error("Partial legacy guardian committed process membership does not match durable recovery state")
|
|
2292
|
+
}
|
|
2293
|
+
|
|
2128
2294
|
/**
|
|
2129
2295
|
* @param {string} candidatePath - Path that must not preexist.
|
|
2130
2296
|
* @param {string} label - Diagnostic label.
|