Skip to content

relay errr #895

Closed
Closed

Description

@kanan-droid

so um iv tested everything from re deploying my relays to re installing the app and certs but nothing works i keep getting relay timeout everything was fine yesterday
im using 10 relays on 10 separate gmails
no i didnt hit the limit
i have tested all of them one by one and they are working fine
in the app logs im just seeing 502 timeout and thats it
plz helpppppp

Activity

  1. Paymanonline commented on May 8, 2026

    @Paymanonline

    There is access issue.
    Delete the project on script.google.com and do it again untill Test relay return "test result: pass".

  2. kanan-droid commented on May 8, 2026

    @kanan-droid
    Author

    @Paymanonline i tried it and still nothing i can open up the relay link on anybrowser and its fine also the auth_keys are matching

  3. therealaleph commented on May 8, 2026

    @therealaleph
    Owner

    Reviewed via Anthropic Claude.

    @kanan-droid — 502 timeout on 10 deployments across 10 separate Gmail accounts when each tested individually works = something Apps-Script-side, not config.

    Most likely cause given the timing: Apps Script edge regional throttle. Google sometimes throttles the script.googleusercontent.com edge serving deployments from a specific region for short windows (minutes to hours). Tests one-at-a-time hit different edges; concurrent use through the proxy hits the same edge and gets rate-limited.

    Diagnostic:

    1. Run with debug logging:
      RUST_LOG=debug mhrv-rs 2>&1 | tee mhrv-debug.log
      
      visit a couple of sites until you see 502s, then post the first 50 lines of mhrv-debug.log.
    2. Specifically look for:
      • relay POST http://<IP>:80/api — what IP is mhrv-rs reaching for Apps Script? Same one across all 502s? If yes, that edge is throttled.
      • Relay failed: ... — what's the actual error string? HTTP 502 from the edge vs io: timeout are very different.

    Workarounds while diagnosing:

    • Set fetch_ips_from_api: true (default) so mhrv-rs rotates Apps Script edge IPs, picking healthy ones.
    • Set parallel_relay: 2 to fan out across 2 deployments per request — sometimes hides flaky-edge behavior.
    • Remove + recreate the deployments. Apps Script "deployment age" sometimes correlates with throttle. Newer deployments often get healthier edges for a few days.
    • Try a different time of day. The throttle has a strong diurnal pattern (worse during Iran peak, ~21:00–01:00 local).

    Will close in 30 days if no log; reopen with debug.log please.

  4. kanan-droid commented on May 8, 2026

    @kanan-droid
    Author

    @therealaleph i think its edge throthling becuz its working fine now evem cloud worker version which is really slow in youtube is now loading youtube fine so yeah

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions