New
- Monitoring endpoints with their own access control (#518, #519). Settings → API can turn on monitoring and set which addresses may read it, and whether each range needs a token. Monitoring tokens are read-only and only open the monitoring endpoints. A new
/metricsendpoint serves the same data in Prometheus format. Monitoring is off by default, and with it off/api/v1/metricsand/api/v1/healthwork as before. Thanks @Glob0x17. - Limit on concurrent offsite syncs (#515). Settings → General has "Max Concurrent Offsite Syncs", default 4. Syncs above the limit wait their turn without holding up backups or prunes. Set 0 for no limit. Thanks @Glob0x17.
Fixes
- Email notifications on port 465 failed when set up from Settings → Apprise (#511). That form saved email URLs without a TLS mode, so Apprise tried STARTTLS on an SSL port. The form now saves the mode, and existing email services get it added. Thanks @mjr6140.
- Repos on extra storage locations were unreachable after a server restore (#514).
bbs-restorenow rebuilds the list of storage paths each client's SSH user may reach. Thanks @Glob0x17. - Login and 2FA attempt limits no longer reset early when another rate-limited endpoint is used.
No agent change; agent stays 2.96.8.