Repository navigation
docs(debug-journal): fold the 2026-11-08..11-10 queue slice; correct the curve, PV calibration and startup-retry entries - #6239
Draft
springfall2008 wants to merge 1 commit into
Draft
springfall2008 wants to merge 1 commit into
springfall2008 wants to merge 1 commit into
Conversation
…the curve, PV calibration and startup-retry entries - fold 12 of the 15 queued candidates, in nine bullets; drop 3 already covered or fixed (6056/#6057, 6072, 6074) - rewrite the charge/discharge curve section for the learned SoC x temperature rate grid (#6212/#6227/#6189) - correct PV calibration to the 28-day evidence store (#6231/#6233), the fatal-error latch (PR #5766 merged) and the balancer (#6194) - record the hardcoded-date failure in test_web_pv_calibration_chart seen on main from 17 Nov Co-Authored-By: Claude Code <noreply@anthropic.com>
This branch has not been deployed
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Automated by the
/journal-updatetriage-bot flow: folds the first 15 queue candidates (plain filename order) intotools/debug-journal.md, re-checking each againstorigin/mainatda2b0a5b, and corrects entries that merges since the last fold have overtaken. 111 candidates remain queued after this slice (126 before it).Candidates folded (12)
write to register 0x56 was not confirmed,stopping powercontrolmode8_trigger ... ['house_load_alt']) are solax-modbus's, not Predbat's. On Predbat not able to control invertor anymore. #6059 bothhouse_load_altand meter 2 were disabled in the entity registry and re-enabling them restored control; docs PR docs(solax): Mode 8 control needs House Load Alt and Meter 2 sensors enabled (#6059) #6070 (merged0c8165c9) covers the user-facing side. Also added theSource: custom_components/...:line→ upstream-tag technique to Versions, logs and attachments.ab5dd9e7), withhistory_checked, the 30-minute recheck less a 15-minute overlap, and thetracked=Falseprecedent. Thefrom_timehalf was already recorded.IPC_TIMEOUT(60 s);engine_accepts_from_time()is the compat-gate precedent.set_charge_window/set_discharge_windowin apps.yaml are dead config, the runtime gates come only from the mode select,has_time_windowis unread, and the charge/discharge midnight-disable writers. Rewrote the candidate's conclusion: on current main the service/immediate paths are gated onset_soc_enable, not the window flags, so Control SOC only does stop the window writes while still driving a script-controlled inverter (except aset_read_only/mode reset). Added the/debug_apps_liveattachment technique to Replays and debug dumps.{power}bullet: since PR Fold balance_inverters into execute: one source of inverter rate writes #5149 the rate is written byapply_rate_intent()afteradjust_*_immediate()builds the payload, so the previous cycle's rate is sent; PR fix: send this cycle's planned rate as {power} to service-driven inverters (#3311, #5252) #5373 is still open (checked today).charger_control_tick()sits inside the 300 sSTATUS_POLL_SECONDSblock as well, so the symptom is a window boundary acted on up to ~5 minutes late on an uptime cadence, not a decision made on stale data.get_from_incrementing()wraps a negative index by +24 h, infetch.py,load_ml_component.pyandpredheat.py.Battery Poweris unsigned, the sign is register 33135; mapbattery_power_net, nobattery_power_invert). PR Updates to solis_local.yaml #6088 is open and main still maps plain_battery_power; marked as verified against the integration source, not hardware.Candidates dropped (3)
0718e5ab) and already covered by the MCP OAuth bullet.limiterTotalPowersince PR Sunsynk Cloud: read grid_power from the CT limiterTotalPower, not pac #6072".03e97def).Existing entries corrected (with the merge that invalidated them)
find_charge_curve()no longer exists: the ceiling is a learned SoC × temperature rate grid (battery_rate_grid_update(),find_rate_grid_samples(),fit_rate_grid(),rate_grid_curve()) from feat(battery): learn charge and discharge rate grids (SoC x battery temperature) #6212/feat(battery): learn the temperature charge curve, keep auto SoC curves in storage and out of the cold #6189, with the site-totalbattery_powerfix (Curve learners: site-wide battery_power (0 on other inverters) is divided by one inverter's max rate - learned rate ~N x too high #6226/fix(curves): learn site-wide battery_power against the whole fleet; save rate grid evidence with an expiry; average battery_temperature_history #6227), the storage guarantees (Battery rate grid store saved without expiry: evidence can expire on storage backends with a default TTL #6225) and the current test names. Removed theif id == 0learner gate,Cannot find battery discharge curve,Sensor history data lengthsand the GH#5219ZeroDivisionError, all overtaken by the rework.days_use_scaling/calibrate()" entry with the 28-day evidence store and shrunk factors from feat(solar): PV calibration from a 28-day evidence store with self-limiting factors #6231/fix(solar): PV calibration follow-ups - restarts, clock changes, cheaper retries, page labels #6233 (3-valid-day minimum, the new log wording, and a note that an old log's wording dates it); added a second bullet for the evidence store's persistence.4c6f7ad2); the entry said it was open. Now records the 5 s / 10 min startup retry and leaves fix(ha): ride out a Home Assistant restart instead of stopping after 10 errors #5449/Fix #5134: don't wait out wait_api_started() for a component whose dependency never constructed #5323 open.min(claimed rate, learned ceiling), andsensor.<prefix>_balance_inverters.test_web_pv_calibration_chartexpectsNext scan 2026-11-17 00:00and fails from today. This is why the quick suite is red on main at the moment; it is not caused by this PR.Consolidation and file size
git log --since="2 weeks ago" origin/mainand each merged PR named above.Verification
./run_pre_commit: every hook passes (cspell, markdownlint, ruff, black, and the repo's own test hooks). The quick suite fails only inweb_pv_calibration_chart(t10 next scan), which fails on clean main from today's date; this change is documentation only and nothing reads it at test time.