Files
coolify/tests/Feature/RemoteProcessTimeoutTest.php
T
Andras Bacsai 599dc1ca67 fix(ssh): do not replay remote commands that failed on the server
The SSH retry matched connection words in any error output. When `mc alias set`
in a backup upload could not reach S3, its "connection refused" output caused a
retry of the full batch, and the second `docker run --name backup-of-<uuid>`
failed with a name conflict that hid the real error.

- Retry only SSH transport failures: exit code 255, or errors without an exit code.
- Keep the remote exit code on deployment command failures.
- Remove a leftover backup helper container in the same upload batch.

Closes #12059
2026-10-03 20:52:01 +02:00

141 lines
4.9 KiB
PHP

<?php
use App\Helpers\SshMultiplexingHelper;
use App\Models\PrivateKey;
use App\Models\Server;
use App\Models\User;
use Illuminate\Foundation\Testing\RefreshDatabase;
use Illuminate\Support\Facades\Process;
use Illuminate\Support\Facades\Storage;
/**
* Tests that per-call SSH command timeouts (e.g. a scheduled backup's configured
* timeout) reach the shell-level `timeout N ssh` wrapper instead of being capped
* by the global constants.ssh.command_timeout default.
*
* @see https://github.com/coollabsio/coolify DatabaseBackupJob/VolumeBackupJob pass
* a per-backup timeout to instant_remote_process()
*/
uses(RefreshDatabase::class);
function makeTimeoutTestServer(): Server
{
$user = User::factory()->create();
$team = $user->teams()->first();
$privateKeyContent = '-----BEGIN OPENSSH PRIVATE KEY-----
'.
'b3BlbnNzaC1rZXktdjEAAAAABG5vbmUAAAAEbm9uZQAAAAAAAAABAAAAMwAAAAtzc2gtZW
'.
'QyNTUxOQAAACBbhpqHhqv6aI67Mj9abM3DVbmcfYhZAhC7ca4d9UCevAAAAJi/QySHv0Mk
'.
'hwAAAAtzc2gtZWQyNTUxOQAAACBbhpqHhqv6aI67Mj9abM3DVbmcfYhZAhC7ca4d9UCevA
'.
'AAAECBQw4jg1WRT2IGHMncCiZhURCts2s24HoDS0thHnnRKVuGmoeGq/pojrsyP1pszcNV
'.
'uZx9iFkCELtxrh31QJ68AAAAEXNhaWxANzZmZjY2ZDJlMmRkAQIDBA==
'.
'-----END OPENSSH PRIVATE KEY-----';
$privateKey = PrivateKey::create([
'name' => 'timeout-test-key-'.uniqid(),
'private_key' => $privateKeyContent,
'team_id' => $team->id,
]);
Storage::fake('ssh-keys');
Storage::disk('ssh-keys')->put("ssh_key@{$privateKey->uuid}", $privateKeyContent);
$server = Server::factory()->create([
'team_id' => $team->id,
'private_key_id' => $privateKey->id,
]);
Storage::disk('ssh-keys')->put("ssh_key@{$server->privateKey->uuid}", $server->privateKey->private_key);
return $server;
}
it('wraps ssh commands with an explicitly passed command timeout', function () {
config(['constants.ssh.mux_enabled' => false]);
$server = makeTimeoutTestServer();
$command = SshMultiplexingHelper::generateSshCommand($server, 'echo ok', commandTimeout: 7200);
expect($command)->toStartWith('timeout 7200 ssh ');
});
it('wraps ssh commands with the configured default timeout when none is passed', function () {
config([
'constants.ssh.mux_enabled' => false,
'constants.ssh.command_timeout' => 1234,
]);
$server = makeTimeoutTestServer();
$command = SshMultiplexingHelper::generateSshCommand($server, 'echo ok');
expect($command)->toStartWith('timeout 1234 ssh ');
});
it('forwards the per-call timeout of instant_remote_process to the ssh timeout wrapper', function () {
config(['constants.ssh.mux_enabled' => false]);
$server = makeTimeoutTestServer();
Process::fake();
instant_remote_process(['echo ok'], $server, timeout: 7200, disableMultiplexing: true);
Process::assertRan(fn ($process) => str_starts_with($process->command, 'timeout 7200 ssh '));
});
it('uses the configured default timeout in instant_remote_process when no timeout is passed', function () {
config([
'constants.ssh.mux_enabled' => false,
'constants.ssh.command_timeout' => 1234,
]);
$server = makeTimeoutTestServer();
Process::fake();
instant_remote_process(['echo ok'], $server, disableMultiplexing: true);
Process::assertRan(fn ($process) => str_starts_with($process->command, 'timeout 1234 ssh '));
});
it('does not replay a command batch when a remote command fails with connection text in its output', function () {
config([
'constants.ssh.mux_enabled' => false,
'constants.ssh.max_retries' => 3,
'constants.ssh.retry_base_delay' => 0,
]);
$server = makeTimeoutTestServer();
Process::fake([
'*' => Process::result(errorOutput: 'mc: <ERROR> Unable to initialize new alias. dial tcp 10.0.0.5:9000: connect: connection refused', exitCode: 1),
]);
expect(fn () => instant_remote_process(['docker run -d --name backup-of-x minio/mc', 'docker exec backup-of-x mc alias set temporary http://s3 k s'], $server, disableMultiplexing: true))
->toThrow(RuntimeException::class, 'connection refused');
Process::assertRanTimes(fn ($process) => str_contains($process->command, 'ssh '), 1);
});
it('retries a command batch when the ssh connection itself fails', function () {
config([
'constants.ssh.mux_enabled' => false,
'constants.ssh.max_retries' => 3,
'constants.ssh.retry_base_delay' => 0,
]);
$server = makeTimeoutTestServer();
Process::fake([
'*' => Process::sequence()
->push(Process::result(errorOutput: 'kex_exchange_identification: read: Connection reset by peer', exitCode: 255))
->push(Process::result(output: 'ok')),
]);
expect(instant_remote_process(['echo ok'], $server, disableMultiplexing: true))->toBe('ok');
Process::assertRanTimes(fn ($process) => str_contains($process->command, 'ssh '), 2);
});