Reconnection error "UnrecognisedName" after standby with Reverse Proxy (WSS)

Software Integrations
1

Hi everyone,

My name is Max and I’m new to this forum. If I’m posting in the wrong section, I would appreciate it if an admin could move this thread to the appropriate place.

I am currently working on an integration driver for IP-Symcon, and the basic functionality is already working well. However, since switching to wss:// with a certificate via my reverse proxy, I’ve noticed that it takes quite a long time for the driver to reconnect after the Remote Two wakes up from standby.

I checked the remote’s logs and noticed the following:

Plaintext

2026-02-28 13:45:34.334550 +00:00  core  NOTICE  [uc_symcon_31544] Resolved visu.your-domain.com: 10.0.40.252
2026-02-28 13:45:34.325432 +00:00  core  NOTICE  [uc_symcon_31544] Session became active: sending subscribe_events, get_driver_version, connect
2026-02-28 13:45:34.324686 +00:00  core  NOTICE  [uc_symcon_31544] SessionAuthenticated msg
2026-02-28 13:45:34.324198 +00:00  core  NOTICE  [uc_symcon_31544] Session state change: Authenticated
2026-02-28 13:45:34.322729 +00:00  core  NOTICE  [uc_symcon_31544] Resolving: visu.your-domain.com
2026-02-28 13:45:34.319758 +00:00  core  NOTICE  [uc_symcon_31544] Connected to: wss://visu.your-domain.com/hook/uc_remote_31544
2026-02-28 13:45:27.380244 +00:00  core  NOTICE  [uc_symcon_31544] Too many failed connection attempts: clearing resolved integration URL to trigger hostname lookup
2026-02-28 13:45:27.379158 +00:00  core  ERROR   [uc_symcon_31544] Error connecting to driver wss://10.0.40.252/hook/uc_remote_31544: ServiceUnavailable("Failed to connect to host: received fatal alert: UnrecognisedName")

It appears that the remote first attempts to establish a connection directly via the IP address. Since my reverse proxy requires the SNI (hostname) to know where to route the request, this attempt fails with UnrecognisedName. Only after several failed attempts and a timeout does the remote seem to trigger a new DNS lookup and successfully connects using the hostname.

Is there a way to force the driver to always use the hostname instead of falling back to the cached IP address?

thx Max

2

You can define a driver_url in the driver metadata instead of using mDNS.

3

I do not use mDNS I used the core API to register the integration. So yes the remote has the correct URL and use it for the first connection and later after the connection fails with the cached IP address.