Plex server suddenly started crashing

Plex Media Server
1

Server: machineIdentifier cf691c3c402f092be642ddea42e7b04d88ba6bda
Platform: Windows Server, Dell PowerEdge T420, build 10.0 (14393)
Current version: 1.43.4.10903-e5521bd8c (was 1.43.3.10861, updated during troubleshooting below)
Plex Pass: Lifetime

Symptom: Server became unable to stay running — library sections stopped loading, playback failed, and the process itself kept exiting within seconds to a couple minutes of launch, with no consistent trigger (happened both at idle and during playback).

What I’ve already tried:

  1. Logs showed repeated SQLITE3 ... 17, statement aborts ... database schema has changed errors. Found the library database’s WAL file had grown to ~365MB — nearly equal to the ~364MB main com.plexapp.plugins.library.db — consistent with an interrupted VACUUM that never checkpointed.
  2. PRAGMA integrity_check via the bundled Plex SQLite.exe came back clean (ok).
  3. Manually ran VACUUM directly against the database (server stopped, outside the main process) — succeeded, WAL cleared, database shrank from 364MB to 358MB.
  4. Updated 1.43.3.10861 → 1.43.4.10903-e5521bd8c (changelog includes PM-5590 timeline-request crash, PM-5627 Tuner Service startup crash, PM-5592, PM-5964, PM-5591, PM-5565).
  5. Did a full clean reinstall (uninstalled via unins000.exe, reinstalled fresh), preserving the existing library database.

Result: each step meaningfully increased uptime before the next crash — seconds → ~80 seconds → ~2.5 minutes — but it still recurs.

No local trace at all: no fatal/error line in Plex Media Server.log (just stops), nothing in Windows Application/System event logs, no WHEA hardware-error events anywhere on the machine, nothing in Defender. Ruled out hardware (no error events ever, and other long-running services on the same box have had zero instability) and ruled out an unrelated IIS install on the same machine (no shared port/HTTP.sys registration).

One crash report did upload automatically via CrashUploader.exe/Sentry, tied to version 1.43.4.10903-e5521bd8c and this account, shortly after the update — hoping someone can help me get real diagnostic traction from that, since local troubleshooting is exhausted.

Happy to attach full logs or anything else that’d help.

2

May want to try the new DBRepair for your DB. I would do a manual db backup before running it, since it’s not final yet.