Bugfix/revert and add more logging for issues #412
Conversation
maslyankov
changed the title
Codecov ReportAttention: Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## main #412 +/- ##
==========================================
- Coverage 71.49% 70.98% -0.52%
==========================================
Files 27 28 +1
Lines 1912 1961 +49
==========================================
+ Hits 1367 1392 +25
- Misses 545 569 +24 ☔ View full report in Codecov by Sentry. |
|
Cannot fix codecov as it gives me "Error 404" when I try to open the results page. It is probably irrelevant as no new functionality has been added to require new tests. |
|
Okay, thanks! Changes are in two of the files only. Other ones are simply pre-commit formatting. |
|
The formatting is a bit of a mess. I pinned ruff in #417 Also a small update the CI script, as my PR tried to publish a new image Can you test these solarman sunsynk changes? |
|
Can you rebase this on the latest? |
…dling and logging - Added logging for suspicious register values (0xFFFF) and large unpacked values. - Improved error handling during unpacking of register values, including specific warnings for single and double register unpacking failures. - Introduced a mapping of register addresses to sensors for better tracking and debugging. - Enhanced performance logging for register reads, including warnings for slow reads and multiple consecutive timeouts. - Updated debug logging to provide more context on register requests and responses. These changes aim to improve the robustness and reliability of the sensor reading process.
maslyankov
force-pushed
the
bugfix/revert-and-attempt-spikes-fix
branch
from
149e306 to
7d5a2d2
Compare
Done |
| raise | ||
|
|
||
| raise ValueError(f"Unsupported number of registers: {len(regs)}") | ||
| except Exception as e: |
There was a problem hiding this comment.
Can we remove this outer general exception handler?
you already log 1&2 register exceptions above
| # Check if state is ok & tracking the sensors being read | ||
| assert self.state is not None | ||
| for sen in sensors: | ||
| if sen not in self.state.values: | ||
| _LOGGER.warning("sensor %s not being tracked", sen.id) | ||
|
|
||
| # Create a map of register addresses to their corresponding sensors | ||
| reg_to_sensor: dict[int, Union[Sensor, None]] = {} |
There was a problem hiding this comment.
| reg_to_sensor: dict[int, Union[Sensor, None]] = {} | |
| reg_to_sensor: dict[int, Sensor | None] = {} |
| isinstance(current_sensor, t) | ||
| for t in ( | ||
| BinarySensor, | ||
| EnumSensor, |
There was a problem hiding this comment.
Can’t 0 be a vlid state for Binary/Enum?
|
I’m ok with the logging during struct unpack, but it seems there are too many logs during read Any chance to rather do this via one of the stats sensors? |
Enhance register unpacking and sensor reading with improved error handling and logging
Revert #408 which caused:
Added logging should help resolve:
Changes:
These changes aim to improve the robustness and reliability of the sensor reading process. This should help us resolve the inconsistency of data when it gets congested.