thrgoodm

Here are diagnostics: phyllis-diagnostics-20260406-1246.zip

My server started crashing randomly after months of stable uptime. The crashes have no obvious trigger and have left nothing useful in syslog. Key symptoms: Server completely frozen — web UI unreachable, shares dropped, Docker containers unresponsive Still pingable after the crash (network stack survived) Nothing in syslog at the crash point — logging just stopped XFS journal recovery running on all 5 array disks on every boot (confirming unclean shutdown each time) Crashes happening on a very consistent 12 hour interval after each reboot System specsUnraid 7.1.4, kernel 6.12.24 31 GiB RAM, no swap configured 5 array disks (2× ST12000NM0127, 2× ST14000VN0008, 1× WD30EFRX) + WD140EDFZ parity 1TB Crucial P310 NVMe cache ~30 Docker containers including Plex, Jellyfin, Sonarr, Radarr, qBittorrent, Bitmagnet, AdGuard 1 VM: Home Assistant OS 13.1 (qcow2) NIC: Realtek RTL8111/8168 (not the problematic RTL8126) What I ruled outRAM — all SMART and memtest clean, no OOM killer events in logs Thermals — all drives under 44°C, no CPU throttling NVMe — 100% available spare, 2% endurance used, zero media errors, never throttled Network driver — RTL8111 is stable and well-supported, not the buggy RTL8126 Disk health — all SMART attributes clean, zero pending/uncorrectable sectors Docker memory leak — RAM available actually increased slightly over 10 hours (though Bitmagnet grew ~340 MiB in 10 hours, worth watching)