A second review of the restart fix found more of the same class. The map uploader's stop() removes its file log handler and start() never added it back, so after one restart it logged nothing to file; start() now restores it. DARC MoWaS set _running before binding its port, so a port in use left it dead but healthy, and werkzeug's make_server calls sys.exit when the port is taken, which got past every except Exception and stopped the bot; start() now binds first and turns that exit into an OSError.
Boot-time starts go through _start_service_at_boot, which records a restart failure when start() raises, so that service also waits out the backoff, and no longer reports a disabled service as a configuration problem. The health-loop step is _restart_unhealthy_services. New tests cover both, the bridges' session cleanup and the map uploader restart; all 13 lifecycle tests fail on dev.
Review found that marking a not-running service inactive for good also retired services that only returned early for a transient reason: packet capture and the map uploader give up after waiting 30 s for the radio, and the bridges return when meshcore is momentarily unset, which is exactly when the health loop restarts them. Such a service now records a restart failure, so the next attempt waits out service_restart_backoff_seconds; a transient failure recovers, a missing configuration costs one attempt per backoff instead of one every 5 s. The restart decision is _service_restart_due, which the tests now exercise directly.
Also from the review: packet capture and the map uploader never cleared the should_exit flag stop() sets, so after a restart their background loops exited at once while the service reported healthy; start() now clears it. The Discord and Telegram bridges close the HTTP session they open when they return early because meshcore is not available.
The health loop restarted every enabled service whose is_healthy() was False. A service that returns from start() without running because it has nothing to do (Discord or Telegram with no channels, the map uploader with no key) is never healthy, and its restart raised nothing, so it counted as a success and cleared the backoff: the service was restarted every 5 seconds, logging a warning each time.
Services whose start() returns without running are now recorded as inactive, logged once, and skipped by the health loop. A restart that leaves a service running but unhealthy records a failure, so the next attempt waits out service_restart_backoff_seconds.
Restarts also double-subscribed: services subscribed to meshcore events in start() and never unsubscribed (the comments said meshcore cleans up; it does not). BaseServicePlugin gains _subscribe and _unsubscribe_all; the Discord, Telegram, map uploader, packet capture and prefix-collision services subscribe through it, unsubscribe in stop(), and drop the old instance's subscription before re-subscribing after a transport reconnect. Tests use meshcore's real EventDispatcher and fail on dev.
- Added `on_transport_reconnected` method to service plugins to re-establish event subscriptions after a transport reconnect.
- Updated `core.py` to notify running services of transport reconnections.
- Enhanced logging for services to track re-subscription actions.
- Added tests to verify the behavior of service plugins during transport reconnects.
- Replaced the use of AdvertFlags enum for flag parsing with bitwise operations to handle invalid values correctly.
- Updated the parsing logic in both MessageHandler and MapUploaderService to use boolean flags for better clarity and performance.
- Adjusted error logging to use warnings instead of errors for parsing issues, enhancing the logging strategy.
- Added unit tests to ensure correct parsing behavior for edge cases with invalid flag values.
- Added configuration options `mqtt_skip_unparseable_packets` and `advert_require_valid_signature` to control MQTT publishing behavior based on packet validity.
- Updated `decode_path_len_byte` function to return `None` for reserved size codes, improving path length validation.
- Implemented logic in `PacketCaptureService` to skip publishing unparseable packets and ADVERT packets with invalid signatures.
- Introduced `verify_meshcore_advert_ed25519` function for signature verification of ADVERT packets, with corresponding unit tests to ensure functionality.
- Enhanced documentation to reflect new configuration options and their effects on packet processing.
- Introduced an asyncio lock to serialize deduplication checks and updates to seen adverts, preventing race conditions during concurrent advert processing.
- Enhanced replay attack detection by ensuring deduplication state is consistent across await boundaries.
- Improved logging for duplicate advert uploads and invalid coordinate handling, providing clearer feedback on advert processing outcomes.
These changes enhance the reliability and integrity of advert uploads in the MapUploaderService.
- Added a new configuration option `condense_paths` in `config.ini.example` to control path formatting.
- Enhanced `MultitestCommand` to condense output paths using shared prefixes and tree branches when `condense_paths` is enabled.
- Introduced utility functions for path processing, including `_condense_path_lines` and `_path_to_tokens`.
- Updated tests to validate the new path condensing functionality and configuration behavior.
BUG-001: web viewer login/session auth (in web viewer commit)
BUG-002: db_manager ALTER TABLE for missing channel_operations and
feed_message_queue columns on startup
BUG-015: scheduler thread blocked on future.result(); replaced all
blocking waits with add_done_callback (fire-and-forget)
BUG-016: reboot_radio sends meshcore.commands.reboot() before disconnect
BUG-017: radio disconnect uses asyncio.wait_for(timeout=10)
BUG-022: custom asyncio loop exception handler suppresses IndexError
from meshcore parser at DEBUG level
BUG-024: last_db_backup_run updated after each run; 2-min startup
window; last-run seeded from DB on restart
BUG-025: send_channel_message retries up to 2 times (2s delay) on
no_event_received via _is_no_event_received() helper
BUG-026: split_text_into_chunks() and get_max_message_length() added
to CommandManager; keyword dispatch uses send_response_chunked()
BUG-028: byte_data = b"" initialised before try block in
decode_meshcore_packet to prevent UnboundLocalError in except handler
TraceCommand: path nodes reversed and return path truncated; fixed
format_elapsed_display: UTC normalisation before elapsed computation (#75)
RepeaterManager: auto_manage_contacts guard before any purge logic (#50)
Command aliases: [Aliases] config section injects shorthands at startup
JSON logging: _JsonFormatter; json_logging = true in [Logging]
Structured JSON logging compatible with Loki, Elasticsearch, Splunk
Discord bridge, Telegram bridge, and all service plugins updated
MeshGraph edge promotion logic corrected
Shutdown: scheduler and meshcore disconnect joined cleanly; log spam fixed
All modules: ruff and mypy cleanup applied (type annotations, imports)
- Introduced a new utility function `decode_path_len_byte` to decode RF packet path length bytes, supporting both legacy and multi-byte paths.
- Updated various modules to utilize the new decoding logic, ensuring compatibility with configured prefix lengths.
- Modified database schemas to include `bytes_per_hop` and `out_bytes_per_hop` columns for better path management.
- Enhanced path parsing and validation across commands and services to accommodate variable prefix lengths.
- Improved logging and error handling for path-related operations, ensuring robustness during transitions.
- Updated logging setup to use default values when the [Logging] section is missing, ensuring consistent behavior.
- Improved handling of colored output and log file configuration based on the presence of the [Logging] section across multiple modules.
- Refactored related code in core and service plugins for better clarity and maintainability.
- Added deep copy of event payloads in multiple modules to avoid segmentation faults when events are freed.
- Updated the handling of payloads in ChannelManager, MessageHandler, DiscordBridgeService, MapUploaderService, and PacketCaptureService to ensure safe access to event data.
- Enhanced logging for cases where payloads are missing, improving error handling and debugging capabilities.
- Replaced the validate_safe_path function with a new resolve_path utility to simplify database path resolution in BotDataViewer, BotIntegration, and MapUploaderService.
- Updated the logic to ensure that both relative and absolute paths are handled correctly, enhancing the robustness of database connections.
- Improved code readability and maintainability by centralizing path resolution logic.
- Centralized placeholder handling in utils instead of individual function handlers
- Updated WxCommand to better handle period identification for tomorrow's weather, ensuring accurate retrieval of daytime and nighttime periods based on current conditions.
- Added a cleanup mechanism in MapUploaderService to manage memory usage by periodically removing old entries from seen_adverts, preventing unbounded growth and improving performance.
- Optimized packet capture service to utilize indexed lookups for faster data retrieval and ensure fallback mechanisms for backward compatibility.
- Clarified comments in config.ini.example and map_uploader_service.py to specify that adverts with any coordinate exactly 0.0 are considered invalid and will be skipped during uploads. This enhances understanding of the coordinate validation logic in the map uploader service.
- Introduced a new [MapUploader] section in config.ini.example to configure the map uploader service, including options for enabling the service, API endpoint, private key path, re-upload interval, and verbose logging.
- Updated requirements.txt to include cryptography and pynacl packages for enhanced security features in the map uploader service.
agessaman