Expiring leases instead of long-held locks
CI / unit (8.2) (push) Has been cancelled
CI / unit (8.3) (push) Has been cancelled
CI / unit (8.4) (push) Has been cancelled
CI / integration (mariadb, 32) (push) Has been cancelled
CI / integration (mariadb, 33) (push) Has been cancelled
CI / integration (mariadb, 34) (push) Has been cancelled
CI / integration (mariadb, 35) (push) Has been cancelled
CI / integration (pgsql, 32) (push) Has been cancelled
CI / integration (pgsql, 33) (push) Has been cancelled
CI / integration (pgsql, 34) (push) Has been cancelled
CI / integration (pgsql, 35) (push) Has been cancelled
CI / integration (sqlite, 32) (push) Has been cancelled
CI / integration (sqlite, 33) (push) Has been cancelled
CI / integration (sqlite, 34) (push) Has been cancelled
CI / integration (sqlite, 35) (push) Has been cancelled

A backup step killed by a time limit, OOM or kill -9 kept Nextcloud's
lock on its location for up to an hour, blocking every new start and the
run itself. Locations and runs are now guarded by leases in ngb_leases
with an expiry: a backup step holds them for its duration plus two
minutes, restores refresh theirs every minute, prune holds 30 minutes,
untrash 5. Nextcloud's lock is only used for the milliseconds in which a
lease is checked and written.

Also: starting a backup while one is already running simply continues
it (no exclusive lease needed), and running to completion waits instead
of spinning while a crashed step's lease has not expired.

Tested on NC34: a backup killed four times at random moments resumes on
its own after the lease expires (about two minutes), and restores
bit-identically.
This commit is contained in:
2026-10-02 14:44:52 +02:00
parent 532b23482e
commit 2610c1ad2f
6 changed files with 174 additions and 42 deletions
+88
View File
@@ -0,0 +1,88 @@
<?php
declare(strict_types=1);
// SPDX-FileCopyrightText: 2026 Dillard Blom
// SPDX-License-Identifier: AGPL-3.0-or-later
namespace OCA\NgBackup\Service;
use OCP\IDBConnection;
use OCP\Lock\ILockingProvider;
use OCP\Lock\LockedException;
/**
* Shared/exclusive leases with an expiry, stored in ngb_leases.
*
* Nextcloud's locks live up to an hour; a PHP process killed while holding one (web cron time
* limit, OOM, kill -9) would block a location that long. A lease expires on its own unless its
* holder keeps refreshing it. Nextcloud's lock is only held for the few milliseconds in which a
* lease is checked and written.
*/
class LeaseService {
public const SHARED = 's';
public const EXCLUSIVE = 'x';
public function __construct(
private IDBConnection $db,
private ILockingProvider $locking,
) {
}
/** @return string lease id; throws LockedException when it conflicts with a live lease */
public function acquire(string $name, string $mode, int $ttl): string {
$guard = 'ng_backup/lease/' . $name;
$this->locking->acquireLock($guard, ILockingProvider::LOCK_EXCLUSIVE);
try {
$now = time();
$qb = $this->db->getQueryBuilder();
$qb->delete('ngb_leases')
->where($qb->expr()->eq('name', $qb->createNamedParameter($name)))
->andWhere($qb->expr()->lt('expires_at', $qb->createNamedParameter($now)));
$qb->executeStatement();
$qb = $this->db->getQueryBuilder();
$qb->select('mode')->from('ngb_leases')->where($qb->expr()->eq('name', $qb->createNamedParameter($name)));
$modes = $qb->executeQuery()->fetchAll(\PDO::FETCH_COLUMN);
if ($modes !== [] && ($mode === self::EXCLUSIVE || in_array(self::EXCLUSIVE, $modes, true))) {
throw new LockedException($name);
}
$id = bin2hex(random_bytes(16));
$qb = $this->db->getQueryBuilder();
$qb->insert('ngb_leases')->values([
'id' => $qb->createNamedParameter($id),
'name' => $qb->createNamedParameter($name),
'mode' => $qb->createNamedParameter($mode),
'expires_at' => $qb->createNamedParameter($now + $ttl),
]);
$qb->executeStatement();
return $id;
} finally {
$this->locking->releaseLock($guard, ILockingProvider::LOCK_EXCLUSIVE);
}
}
/** Extend a lease while its holder is still working. */
public function refresh(string $id, int $ttl): void {
$qb = $this->db->getQueryBuilder();
$qb->update('ngb_leases')->set('expires_at', $qb->createNamedParameter(time() + $ttl))
->where($qb->expr()->eq('id', $qb->createNamedParameter($id)));
$qb->executeStatement();
}
public function release(string $id): void {
$qb = $this->db->getQueryBuilder();
$qb->delete('ngb_leases')->where($qb->expr()->eq('id', $qb->createNamedParameter($id)));
$qb->executeStatement();
}
/** Run $fn under a lease; $refresh() can be called during long work. */
public function with(string $name, string $mode, int $ttl, callable $fn): mixed {
$id = $this->acquire($name, $mode, $ttl);
try {
return $fn(fn () => $this->refresh($id, $ttl));
} finally {
$this->release($id);
}
}
}