Hi Railway community,

I'm experiencing sporadic high P99 latency spikes (10-20 seconds) on my Next.js 15 application hosted on Railway Metal in EU West Amsterdam.

Context:

- Framework: Next.js 15 + React 19

- Region: EU West Amsterdam (both app and all 4 PostgreSQL databases)

- Serverless: Disabled (always running)

- Frequency: Very rare - only 1-2 times in the past 15 days

- Impact: When it happens, the entire site becomes slow for that request

What I've investigated:

- Serverless is OFF (no cold starts)

- Database connections are fast (~350ms each)

- Data is small (419 leads, 231KB total JSON)

- All services in same region (no cross-regional latency)

- No correlation with deployments

Questions:

1. Are there any known sporadic latency issues with Railway Metal in EU West Amsterdam?

2. Could this be related to PostgreSQL maintenance (VACUUM/ANALYZE)?

3. Is there a way to see infrastructure-level logs for these specific timestamps?

Any insights would be appreciated!