What best describes your playback issue?
· Music doesn't start when I press "Play"
What type of Zone is affected by this problem?
· *Network Zones* are affected.
Is the affected network Zone connected with Ethernet or WiFi?
· WiFi
Does the issue affect all file formats?
· The issue affects *multiple/all* file formats.
Does the issue happen with local library music, streaming service music, or both?
· *Only streaming* music is affected.
Please select the streaming service(s) with which you're encountering playback problems.
· TIDAL
Have you tried logging out and back in again to your streaming service in Roon Settings?
· Logging out and back in had no impact, the issue remains
Do you have an approximate timestamp of when the issue last occurred?
· Most recent: Aug 5 2026, 12:31 PM Eastern (16:31:04-16:31:15 UTC) - "If Love Is A Skill", "Changes", "Oh! You Pretty Things", and "Feel" all failed at 0% in a row. Earlier same day: 7:32-7:33 AM Eastern (11:32:32-11:33:27 UTC) - "Is This Love" (TIDAL 277080645) and "If Love Is A Skill". Full UTC windows with matching swim/session/begin latencies are in the description.
What are the make and model of the affected audio device(s) and the connection type?
· iPhone running the current Roon ARC app. In the car: CarPlay to the vehicle head unit. At home: phone output directly. Only ARC is affected - all home RAAT zones play from the same core without issues. The failures happen in the ARC app itself regardless of whether CarPlay is attached.
Describe the issue
At the start of every fresh Roon ARC session (iPhone 16 Pro, usually CarPlay in the car, but it also happens at home on Wi-Fi), the first 1-5 play attempts of TIDAL tracks fail immediately with "an unexpected error occurred". Skipping to another track and retrying eventually works, and once the first track plays, the rest of the session is flawless (every later track scrobbles at 100%).
Server logs show the failures line up exactly with the core's call to POST https://api.roonlabs.net/swim/1/session/begin?tidal=max taking 10-18 SECONDS when a new streaming session starts. The same endpoint returns in 80-160 ms for periodic background calls, and other api.roonlabs.net calls made in the same seconds (e.g. /metadata/1/tracks/translate) return in under 200 ms, so this looks specific to the swim session-begin path on the cloud side, not my network.
Evidence from RoonServer_log.txt (timestamps UTC):
08/04 19:54:59 [easyhttp] POST swim/1/session/begin?tidal=max returned after 10992 ms, status 200
08/04 22:30:20 [easyhttp] POST swim/1/session/begin?tidal=max returned after 18172 ms, status 200
08/05 11:36:46 [easyhttp] POST swim/1/session/begin?tidal=max returned after 11256 ms, status 200
Correlation: on 08/04, "Make Me Feel" failed five times at 0s/186s (0%) between 22:26:43 and 22:27:27 UTC; the 18.2 s session/begin completed at 22:30:20 and the same track then played to 100% at 22:30:32. On 08/05, "Is This Love" (tidal track 277080645) failed twice and "If Love Is A Skill" failed at 0% between 11:32:32 and 11:33:27 UTC; after the 11.3 s session/begin completed at 11:36:57, playback worked for the rest of the drive. On 08/05 at 16:31:04-16:31:15 UTC, four tracks failed in a row at 0% while the core was tearing down the previous expired swim session.
Possibly related, recurring every 4 hours: "couldn't get keypair 0d808b4d-...: Result[Status=NotFound]" (HTTP 404 from roonmobile.roonlabs.net), which looks like a stale ARC keypair registration.
Setup: Roon Server 2.71 (build 1680) production on Ubuntu 26.04 LTS, Intel NUC, 16 GB RAM; server healthy during all incidents (1.8 GB RSS, no swap, load 0.05). Library 8,721 tracks. TIDAL (Max) and Qobuz. ARC connects through a WireGuard relay on AWS because my fiber ISP uses CGNAT; the relay is verified NOT the issue, since during the failure windows ARC API requests (/pages/home, /sync) were served in under 200 ms over the same path, and the core's outbound HTTPS to api.roonlabs.net goes directly out the fiber line, not through the tunnel. Core ID: 97c0aa74-7c44-45e0-8e27-614303a48467.
Request: please check server-side latency on swim/1/session/begin for TIDAL session starts. The UTC windows above should be easy to find on your side; happy to enable diagnostics if needed.
Describe your network setup
FairlawnGig fiber ISP (CGNAT, no public IPv4) -> Ubiquiti UCG-Fiber router (192.168.20.1, DHCP/DNS) -> Intel NUC Roon Server wired via gigabit Ethernet (192.168.20.10). Because of CGNAT, ARC port 55000 is forwarded through a WireGuard tunnel from the NUC to an AWS EC2 relay instance (Ohio); ARC connectivity test reports Success with the relay's public IP. Phone uses Verizon LTE in the car or home Wi-Fi (same UCG-Fiber). Note: the ARC data path through the relay is fast and verified working during the failure windows; the slow calls are the core's own outbound HTTPS to api.roonlabs.net, which goes directly out the fiber line, not through the tunnel.