Description of the issue you're facing:

My MongoDB service on Railway is crashing repeatedly in a loop. The service starts up, WiredTiger storage engine opens, then immediately crashes with fatal assertion errors. This creates an infinite restart cycle where the database never becomes available.

Error messages and descriptions:

- ***aborting after fassert() failure

- Fatal assertion

- WiredTiger error message (repeated multiple times)

- Detected unclean shutdown - Lock file is not empty

- WiredTiger recoveryTimestamp errors

Logs (build and/or deploy):

The crash pattern repeats every 2-3 seconds:

1. MongoDB starts successfully

2. WiredTiger opens and attempts recovery

3. Fatal assertion occurs

4. Service aborts and restarts

5. Cycle repeats indefinitely

Resource Usage Analysis (Hobby Plan):

- Current Plan: Hobby Plan

- Included Resources: 8 GB RAM, 8 vCPU, 100 GB Shared Disk

- Current Storage Usage: ~100 MB out of 500 MB available

- Resource Utilization: Well within Hobby Plan limits

- Issue: This is NOT a resource limitation problem

Technical Assessment:

This appears to be WiredTiger storage engine corruption, not a resource constraint issue. The database crashes immediately upon startup before any significant resource usage occurs. The fatal assertion errors indicate storage engine file corruption rather than memory or CPU limitations.

Link to GitHub repo/code or template you're using, if applicable:

N/A - This is a database service issue, not code-related.

Additional Information:

- Service: MongoDB

- Platform: Railway

- Issue started: September 20, 2025

- Impact: Complete database unavailability

- No recent changes made to database configuration

- Resource usage is minimal and well within plan limits