Full restore over SFTP: read packs ahead on four connections at once

A restore read one pack at a time over one SFTP connection, which over a long
line is bound by the round trip and by what one TCP connection gets: about
2-3 MiB/s from a Hetzner Storage Box to Canada, with PHP mostly waiting.

Repository::restore() now first lists the ranges of packs the restore needs,
in order (ReadAheadPlan), and the location reads them ahead into unlinked
files in a spool folder in the target (IReadsAhead). Over SFTP,
SftpRangeFetcher does that on four connections of its own, 16 requests in
flight each, taking the answers of whichever connection has one. A range that
fails, a damaged copy, or a blob that moved since the plan is read the usual
way, which reports what is wrong. Other locations restore as before.

On a test line with 120 ms RTT and 0.2% loss, a 1.6 GiB restore went from
5.1 to 17.8 MiB/s, all files identical.
This commit is contained in:
2026-10-19 01:40:49 +02:00
parent b789d17636
commit ccbbf01c54
12 changed files with 1228 additions and 5 deletions
+46 -1
View File
@@ -14,7 +14,7 @@ use OCP\Files\Storage\IWriteStreamStorage;
* IBackend on top of a Nextcloud storage object (from files_external or core), below a base
* directory on that storage.
*/
final class StorageBackend implements IBackend {
final class StorageBackend implements IBackend, IReadsAhead {
/** Object stores write whole objects atomically; other storages get a temp name + rename. */
private bool $atomicPut;
/** SFTP: read through SftpReadStream, which keeps several requests in flight. */
@@ -243,6 +243,51 @@ final class StorageBackend implements IBackend {
}
}
/**
* Over SFTP: SftpRangeFetcher, on CONNECTIONS new connections of its own (built with the
* reconnect function), so the storage's connection stays free for other requests. Elsewhere
* not (yet): null.
*/
public function canReadAhead(): bool {
return $this->sftpReads && $this->reconnect !== null;
}
public function readAhead(array $ranges, string $spool): ?\Generator {
$reconnect = $this->reconnect;
if (!$this->sftpReads || $reconnect === null || $ranges === []) {
return null;
}
$storages = [];
$connections = [];
$root = '';
for ($n = 0; $n < SftpRangeFetcher::CONNECTIONS; $n++) {
try {
$storage = $reconnect();
$sftp = SftpReadStream::connection($storage);
if ($sftp === null || !is_callable([$storage, 'getRoot'])) {
break;
}
$root = (string)$storage->getRoot();
} catch (\Throwable) {
break; // the server may allow fewer connections: go on with those that work
}
$storages[] = $storage;
$connections[] = $sftp;
}
if ($connections === []) {
return null;
}
$abs = array_map(fn (array $r): array => [$this->abs($r[0]), $r[1], $r[2]], $ranges);
return (static function () use ($storages, $connections, $root, $abs, $spool): \Generator {
// the storages hold their connections: kept until the fetcher is done
try {
yield from SftpRangeFetcher::fetch($connections, $root, $abs, $spool);
} finally {
unset($storages);
}
})();
}
public function getRange(string $path, int $offset, int $length): string {
return $this->call(fn () => $this->readRange($path, $offset, $length));
}