Repository navigation
Conversation
The outage start is stamped by the esp-mqtt or WiFi event task, so it can be a few ms newer than the main loop's now. The unsigned subtraction then wrapped to ~2^32 ms, passed the threshold at once, and sent "down 1193h2m" for a blip that recovered seconds later.
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Problem
A ThinkNode M7 on the current dev release sent "MQTT slot 5 (analyzer-us) down 1193h2m" followed in the same minute by "recovered after 0m". 1193h2m is 2^32 ms.
The alerter takes
nowon the main loop, then reads the outage start, which the esp-mqtt task (or the WiFi event task) stamps with its ownmillis(). A disconnect landing in between leaves the start newer thannow, the unsigned subtraction wraps, and the result passes the "down longer than N minutes" threshold at once.Fix
AlertFaultPolicy::downDurationMstreats a start up to 60 s newer thannowas zero elapsed. The MQTT slot alerts and the WiFi/Ethernet alert both go through it, on repeater and room server.Side effect: an outage whose age is in the last 60 s before the 49.7-day wrap reads as zero for that minute.
Testing
pio test -e native -f test_alert_fault_policy: 26 of 26 pass, including two new cases (start stamped afternow, and the same across themillis()rollover).