kutesir/Kisani
1
0
Fork 0
Code Issues 4 Pull Requests Actions Packages Projects Releases Wiki Activity

fix: resolve duplicate-file failures, alreadySafe reset, and backup restart loop

Browse Source 
BackupEngine: after all retries fail, call fileExists() before recording a
failure. If the file is on NAS (uploaded earlier but fileExists silently
threw on the pre-check), count it as skipped and add it to manifestEntries —
this ensures writeManifest includes every safe asset and prevents alreadySafe
from dropping to 1 after the post-backup reconcile.

BackupView: remove the immediate refreshAndWait from onChange(.failed). That
call raced with refreshAfterBackup's writeManifest, cancelled the manifest-
correct refresh, then reconciled against a stale 1-entry NAS manifest.
Auto-resolve (.failed → .completed when needBackup == 0) is now owned by the
coordinator so it runs after writeManifest finishes.

AutoBackupCoordinator: add backupTriggerArmed flag (set by onActive/
handleNASReachable, cleared by handleReconciliationComplete). Post-backup
reconciles from refreshAfterBackup are no longer able to re-trigger auto-
backup — only an explicit app-open or LAN-join can arm the trigger. Also
added auto-resolve: calls engine.resolveSuccess() when job is .failed but
needBackup == 0, which fires the .completed green-flash path in BackupView.

Co-Authored-By: Kutesir <kutesir@provoc.ug>
Co-Authored-By: Sentry <sentry@provoc.ug>
This commit is contained in:
Robin Kutesa
2026-05-18 18:55:27 +03:00
parent 0acf111d32
commit 8d8b6c4418
3 changed files with 70 additions and 25 deletions
Download Patch File Download Diff File Expand all files Collapse all files

15
Features/Backup/BackupView.swift
View File

@@ -153,18 +153,11 @@ struct BackupView: View {
try? await Task.sleep(nanoseconds: 500_000_000)
statusService.refresh(force: true)
}
} else if case .failed = status {
// Reconcile — if all files landed on NAS despite the failure, auto-resolve to .completed
Task {
await statusService.refreshAndWait(force: true)
if statusService.snapshot.needBackup == 0 {
engine.resolveSuccess()
withAnimation(.spring(response: 0.3, dampingFraction: 0.8)) { ctaFlashGreen = true }
try? await Task.sleep(nanoseconds: 2_500_000_000)
withAnimation(.spring(response: 0.5, dampingFraction: 0.8)) { ctaFlashGreen = false }
}
}
}
// .failed auto-resolve is owned by AutoBackupCoordinator.handleReconciliationComplete.
// It waits for refreshAfterBackup's writeManifest to finish before reconciling, then
// calls engine.resolveSuccess() when needBackup == 0, which re-triggers .completed above.
// Calling refreshAndWait here would race with writeManifest and read a stale NAS manifest.
}
}

37
Services/AutoBackupCoordinator.swift
View File

@@ -33,6 +33,11 @@ final class AutoBackupCoordinator: ObservableObject {
private var lastAutoCheckAt: Date = .distantPast
private var recheckTimer: AnyCancellable?
// One-shot flag armed by onActive()/handleNASReachable() and consumed by
// handleReconciliationComplete(). Prevents the post-backup refresh cycle from
// re-triggering auto-backup — only an explicit app-open or LAN-join should do that.
private var backupTriggerArmed = false
private let engine = BackupEngine.shared
private let statusService = BackupStatusService.shared
private let store = ConnectionStore.shared
@@ -62,6 +67,7 @@ final class AutoBackupCoordinator: ObservableObject {
Task { @MainActor [weak self] in
self?.recheckTimer = nil
self?.lastAutoCheckAt = .distantPast
self?.backupTriggerArmed = false
log.debug("App resigned — recheck timer stopped, debounce reset")
}
}
@@ -120,6 +126,7 @@ final class AutoBackupCoordinator: ObservableObject {
}
lastAutoCheckAt = Date()
backupTriggerArmed = true
phase = .checking
startPeriodicRecheckIfNeeded()
log.info("onActive: trigger=\(lanTriggered ? "LAN" : "appOpen") — status check starting")
@@ -134,13 +141,25 @@ final class AutoBackupCoordinator: ObservableObject {
// No destination configured yet
guard let conn = store.savedConnection, conn.remotePath != "/" else {
phase = .idle
backupTriggerArmed = false
return
}
// Nothing to back up — clear any stale state
// Auto-resolve: engine reported failure but reconcile shows all assets are safe.
// This fires after refreshAfterBackup's writeManifest completes and updates the manifest,
// at which point needBackup correctly reflects what's still missing.
if case .failed = engine.job.status {
if snapshot.needBackup == 0 {
engine.resolveSuccess()
log.info("Auto-resolved .failed → .completed (needBackup == 0 after reconcile)")
}
}
// Nothing to back up
guard snapshot.needBackup > 0 else {
phase = .idle
pendingAtDisconnect = 0
backupTriggerArmed = false
return
}
@@ -148,23 +167,35 @@ final class AutoBackupCoordinator: ObservableObject {
if snapshot.connectionState == .offline {
pendingAtDisconnect = snapshot.needBackup
phase = .disconnected
backupTriggerArmed = false
log.info("NAS offline — \(snapshot.needBackup) items queued")
return
}
// Only fire auto-backup when explicitly armed by onActive() or handleNASReachable().
// Post-backup reconciles (from refreshAfterBackup) must not re-trigger — that's the
// source of the restart loop when alreadySafe is temporarily wrong.
guard backupTriggerArmed else {
phase = .idle
log.info("handleReconciliationComplete: trigger not armed — skipping auto-backup")
return
}
// Gate auto-backup on trigger source:
// LAN join → autoBackupEnabled (lanTriggerEnabled guards the SSID check in LANMonitor)
// App open → autoBackupOnOpen (off by default; requires explicit opt-in)
// LAN join → autoBackupEnabled
// App open → autoBackupOnOpen (off by default)
let autoAllowed = lastTriggerWasLAN
? store.autoBackupEnabled
: store.autoBackupOnOpen
guard autoAllowed else {
phase = .idle
backupTriggerArmed = false
log.info("Auto-backup gated — trigger=\(self.lastTriggerWasLAN ? "LAN" : "appOpen") enabled=\(autoAllowed)")
return
}
backupTriggerArmed = false
triggerAutoBackupIfNeeded(conn: conn)
}

43
Services/BackupEngine.swift
View File

@@ -306,17 +306,38 @@ final class BackupEngine: ObservableObject {
}
if !uploadSucceeded, let error = lastUploadError {
logger.error("Upload permanently failed for \(asset.filename, privacy: .public): \(error.localizedDescription, privacy: .public)")
job.fileFailed()
failed += 1
if assetIdx < queueItems.count {
queueItems[assetIdx].status = .failed
queueItems[assetIdx].completedAt = Date()
queueItems[assetIdx].error = BackupQueueItem.UploadError(
code: (error as NSError).domain,
message: error.localizedDescription,
timestamp: Date()
)
// Before recording a failure, verify the file isn't already on NAS.
// fileExists() may have returned nil earlier due to a transient SMB error;
// if it's actually there, the asset is safe — count it as skipped, not failed.
if (try? await transfer.fileExists(at: remotePath)) == true {
job.fileCompleted(skipped: true)
skipped += 1
if assetIdx < queueItems.count {
queueItems[assetIdx].status = .skipped
queueItems[assetIdx].completedAt = Date()
}
manifestEntries.append(ManifestEntry(
localIdentifier: asset.localIdentifier,
filename: asset.filename,
creationDate: asset.creationDate,
fileSize: 0,
remotePath: remotePath,
uploadedAt: Date()
))
logger.info("Upload failed but file confirmed on NAS — marking safe: \(asset.filename, privacy: .public)")
} else {
logger.error("Upload permanently failed for \(asset.filename, privacy: .public): \(error.localizedDescription, privacy: .public)")
job.fileFailed()
failed += 1
if assetIdx < queueItems.count {
queueItems[assetIdx].status = .failed
queueItems[assetIdx].completedAt = Date()
queueItems[assetIdx].error = BackupQueueItem.UploadError(
code: (error as NSError).domain,
message: error.localizedDescription,
timestamp: Date()
)
}
}
}
}