Repository navigation
[huawei-sun2000] Driver issue on 1.0.0 #72
Description
Device make, model and firmware
192.168.68.123
Sanitized FTW driver diagnostics
## FTW driver diagnostics
- Driver: `huawei-sun2000`
- Driver version: `1.0.0`
- Source: `bundled`
- FTW version: `v1.10.5`
- Host API: `1`
- Runtime ABI: `gopher-lua-source-v1`
- Device make: `Huawei`
- Health: `offline`
- Last error: present; omitted here for privacy
## Test steps
1. Opened the driver Diagnose view.
2. Checked current health and telemetry.
3. Add the steps that reproduce the issue, without IP addresses, serial numbers, credentials, site IDs, raw config, or logs.What happened?
Can’t connect to the unit. I know Sourseful had some problems with this to. Think it was that the timeout was Done to early. It work with Sourceful
Steps to reproduce
No connection
Expected result
Not Connecting
Privacy check
- I removed IP addresses, serial numbers, credentials, site IDs, raw config, logs and private documents.
Activity
frahlg commentedon Aug 5, 2026
Thanks for reporting this. A couple of things worth flagging before this is actionable:
The diagnostics show a very old driver.
Driver version: 1.0.0,Source: bundled— but the currenthuaweicatalog driver in this repo is at 2.1.1, with several read-stability fixes since 1.0.0 (bounded register probing so a register that doesn't answer stops being retried instead of failing every poll — seedrivers/tests/test_absent_register_settles.pyand the CHANGELOG).bundledmeans this device is reading from FTW's local recovery snapshot rather than the live catalog channel, and that snapshot only moves when someone re-pinsBUNDLED_SOURCE.jsonon the FTW side — this repository merging a fix doesn't reach it automatically (seeAGENTS.md, "Where a driver change has to land"). So it's possible this is already fixed upstream and just hasn't reached this installation.The report itself doesn't have enough to reproduce. "Can't connect", "no connection", and a redacted last error don't point at a specific register, timeout value, or failure mode — and "Sourceful had some problems with this too... It work with Sourceful" is ambiguous about what's being compared to what. To actually look into this, it would help to know:
Last errortext from the Diagnose view — is it a TCP connect timeout, a connection refused, or a Modbus exception on a specific register?Not merging or changing anything here — this needs more detail from the reporter before it's a driver bug rather than a config/connectivity question.
Generated by Claude Code
frahlg commentedon Aug 7, 2026
Nu finns en konkret retest-build:
Uppdatera till v1.16.1-beta.10 och kör Diagnose igen. Om den fortfarande blir offline, lägg in den exakta men sanerade Last error-texten samt om port 502 går att nå och om enheten kör embedded Modbus TCP eller SDongle. Då vet vi om 2.1.1 löste den gamla read/timeout-vägen eller om detta är ett separat anslutningsfel.
Jag lämnar issuet öppet tills det finns ett resultat från hårdvaran.
goranklang commentedon Aug 7, 2026
frahlg commentedon Aug 7, 2026