OpenData Buffer: HA pipelines without Kafka
apurvamehta
12 points
1 comment
April 30, 2026
Related Discussions
Found 5 related stories in 94.4ms across 8,303 title embeddings via pgvector HNSW
- OpenData Timeseries: Prometheus-compatible metrics on object storage apurvamehta · 13 pts · April 16, 2026 · 52% similar
- OpenData Vector: MIT-Licensed Vector Search on Object Storage apurvamehta · 35 pts · May 14, 2026 · 49% similar
- Show HN: IceGate – Observability data lake engine mineev · 14 pts · April 13, 2026 · 48% similar
- Moving a large-scale metrics pipeline from StatsD to OpenTelemetry / Prometheus jmarbach · 73 pts · April 16, 2026 · 48% similar
- Honker – Durable queues, streams, pub/sub, and cron scheduler in a SQLite file ferriswil · 200 pts · April 30, 2026 · 46% similar
Discussion Highlights (1 comments)
agavra
Save you the read, here's the interesting part: For $90/mo, we ran a benchmark pushing ~30 MiB/s of metrics data via Buffer. Using public pricing calculators, the equivalent Kafka service cost roughly $1,300/month for managed WarpStream and $8,500/month for self-hosted Apache Kafka.