Files
artifactapi/internal/database/github_sync.go
T
unkin-agent 492607a164
ci/woodpecker/tag/docker Pipeline was successful
Retry failed GitHub release scans with backoff (#133)
Releasing a GitHub sync lease always advanced `last_synced_at`, even after a failed scan (e.g. a rate-limit 403). A remote that failed once waited a full `mutable_ttl` before retrying, and a cold remote kept returning 503 until then.

- record scan outcomes in one shared lease helper for github_rpm/deb/alpine
- keep `last_synced_at` and the ETag on failure; retry from 60s with exponential backoff, capped at min(10m, ttl/4)
- honour `Retry-After` / `X-RateLimit-Reset`, clamped to `mutable_ttl`
- add `sync_failures` / `next_retry_at` columns (migration 0002)

Reviewed-on: #133
Co-authored-by: unkin-agent <unkin-agent@unkin.net>
Co-committed-by: unkin-agent <unkin-agent@unkin.net>
2026-10-09 23:15:55 +11:00

102 lines
4.2 KiB
Go

package database
import (
"context"
"errors"
"time"
"github.com/jackc/pgx/v5"
"git.unkin.net/unkin/artifactapi/internal/provider"
"git.unkin.net/unkin/artifactapi/pkg/models"
)
// ListGitHubRPMRemotes returns every github_rpm remote so the syncer can sweep
// them on each poll tick.
func (db *DB) ListGitHubRPMRemotes(ctx context.Context) ([]models.Remote, error) {
rows, err := db.Pool.Query(ctx, `SELECT `+remoteCols+` FROM remotes WHERE package_type = $1 ORDER BY name`, models.PackageGitHubRPM)
if err != nil {
return nil, err
}
defer rows.Close()
var remotes []models.Remote
for rows.Next() {
var r models.Remote
if err := scanRemote(rows, &r); err != nil {
return nil, err
}
remotes = append(remotes, r)
}
return remotes, rows.Err()
}
// ClaimGitHubSyncLease atomically claims the per-remote github_rpm sync lease.
// See claimSyncLease.
func (db *DB) ClaimGitHubSyncLease(ctx context.Context, remoteName, owner string, freshness, lease time.Duration) (bool, string, error) {
return db.claimSyncLease(ctx, "github_rpm_sync_state", remoteName, owner, freshness, lease)
}
// ReleaseGitHubSyncLease records a github_rpm scan outcome and frees the lease.
// See releaseSyncLease.
func (db *DB) ReleaseGitHubSyncLease(ctx context.Context, remoteName, owner string, res provider.SyncResult) error {
return db.releaseSyncLease(ctx, "github_rpm_sync_state", remoteName, owner, res)
}
// claimSyncLease atomically claims a per-remote sync lease in table. It succeeds
// (claimed=true) only when the remote is due and no live lease is held by
// another replica. Due means: a pending retry after a failed scan has reached
// next_retry_at, or, with no retry pending, the remote was never synced or was
// synced longer than freshness ago. A zero freshness (prime scans) ignores the
// recency gate but still honours a pending retry's backoff. The returned etag is
// the stored releases-list ETag, shared across replicas so a conditional request
// can short-circuit an unchanged repo.
func (db *DB) claimSyncLease(ctx context.Context, table, remoteName, owner string, freshness, lease time.Duration) (bool, string, error) {
row := db.Pool.QueryRow(ctx, `
INSERT INTO `+table+` AS s (remote_name, sync_lease_owner, sync_lease_expires)
VALUES ($1, $2, now() + make_interval(secs => $4))
ON CONFLICT (remote_name) DO UPDATE
SET sync_lease_owner = $2,
sync_lease_expires = now() + make_interval(secs => $4)
WHERE (CASE WHEN s.next_retry_at IS NOT NULL THEN s.next_retry_at <= now()
ELSE s.last_synced_at IS NULL OR s.last_synced_at < now() - make_interval(secs => $3) END)
AND (s.sync_lease_expires IS NULL OR s.sync_lease_expires < now())
RETURNING s.etag
`, remoteName, owner, freshness.Seconds(), lease.Seconds())
var etag string
if err := row.Scan(&etag); err != nil {
if errors.Is(err, pgx.ErrNoRows) {
return false, "", nil
}
return false, "", err
}
return true, etag, nil
}
// releaseSyncLease records a scan outcome and frees the lease; only the owning
// replica may release. Success advances last_synced_at, persists the etag and
// clears any retry. Failure leaves last_synced_at and etag untouched (the last
// good metadata keeps serving) and schedules next_retry_at with exponential
// backoff from the consecutive-failure count, never before res.RetryAt.
func (db *DB) releaseSyncLease(ctx context.Context, table, remoteName, owner string, res provider.SyncResult) error {
var retryAt *time.Time
if !res.RetryAt.IsZero() {
retryAt = &res.RetryAt
}
_, err := db.Pool.Exec(ctx, `
UPDATE `+table+`
SET last_synced_at = CASE WHEN $3 THEN last_synced_at ELSE now() END,
etag = CASE WHEN $3 THEN etag ELSE $4 END,
sync_failures = CASE WHEN $3 THEN sync_failures + 1 ELSE 0 END,
next_retry_at = CASE WHEN $3 THEN GREATEST(
now() + make_interval(secs => LEAST($5 * power(2, LEAST(sync_failures, 20)), $6)),
$7::timestamptz)
END,
sync_lease_owner = '', sync_lease_expires = NULL
WHERE remote_name = $1 AND sync_lease_owner = $2
`, remoteName, owner, res.Failed, res.Etag, res.Backoff.Seconds(), res.MaxBackoff.Seconds(), retryAt)
return err
}