ProxMenux v1.1.9.4 (Beta)
This release adds Telegram Topics support, visual update indicators, Groq connection improvements, and continued performance optimizations.
Main changes in v1.1.9.4
Telegram Topics Support
Notifications can now be sent to specific topics within Telegram groups that have Topics enabled.
- Topic ID Configuration: New optional field to specify which topic receives notifications
- Automatic Detection: System detects if your group uses topics
- Backwards Compatible: Groups without topics continue working as before
To find your Topic ID, check the URL when viewing a topic: https://t.me/c/XXXXX/123 - the last number (123) is the Topic ID.
Dashboard Update Indicator
The ProxMenux logo now visually indicates when updates are available.
- Dynamic Logo: Logo changes to update variant when new versions are detected
- Non-Intrusive: No popups or banners - just a subtle visual cue
- Dual Channel: Indicates updates for both stable and beta versions
- Persistent State: Status stored in config.json, reset by update scripts
Groq Provider Fix
Resolved connection issues with the Groq AI provider.
- Fixed JSON Parsing: Corrected response handling for Groq's API format
- Improved Error Handling: Better timeout and connection error management
- Model Compatibility: Ensured compatibility with latest Groq model endpoints
Performance Optimizations
Additional improvements to reduce system resource usage.
Optimized API Response Caching
| Resource | Cache TTL | Benefit |
|---|---|---|
| Cluster resources | 30s | Reduced API calls from 6 to 1 per request |
| Sensor readings | 10s | Prevents redundant temperature queries |
| SMART data | 30 min | Disk health cached between polls |
| Hardware info (lspci) | 5 min | Static data cached appropriately |
GitHub Actions Updated
- Upgraded all workflow actions to v6 for Node.js 24 compatibility
- Eliminates deprecation warnings in CI/CD pipeline
Disk Detection Improvements
Enhanced parsing of SMART data for drives with non-standard value encoding.
- Power On Hours Validation: Detects and corrects absurdly large values (billions of hours)
- Intelligent Bit Masking: Extracts correct data from drives that encode extra info in high bytes
- Graceful Fallback: Shows "N/A" instead of impossible values when data cannot be parsed