A lightweight, ultra-fast tool for building observability pipelines

A lightweight, ultra-fast tool for building observability pipelines

https://vector.dev/

Take control of your observability data

Collect, transform, and route all your logs and metrics with one simple tool.

---

Why Vector?

Ultra fast and reliable

Built in Rust, Vector is blistering fast, memory efficient, and designed to handle the most demanding workloads.

End to end

Vector strives to be the only tool you need to get observability data from A to B, deploying as a daemon, sidecar, or aggregator.

Unified

Vector supports logs and metrics, making it easy to collect and process all your observability data.

Vendor neutral

Vector doesn’t favor any specific vendor platforms and fosters a fair, open ecosystem with your best interests in mind. Lock-in free and future proof.

Programmable transforms

Vector’s highly configurable transforms give you the full power of programmable runtimes. Handle complex use cases without limitation.

Clear guarantees

Guarantees matter, and Vector is clear on which guarantees it provides, helping you make the appropriate trade-offs for your use case.

---

https://wiki.eryajf.net/pages/0322lius/#_1-2-2-1-remap

0.2 為什麼用 vector

如果說官方有自賣自誇之嫌，那麼我給出自身實踐的資料以供參考：

本人所在公司每天產生約 15T 的日誌量，在公司日誌架構中 logstash 起著這樣的作用：從 kafka 中消費資料，然後進行清洗、格式轉換，最終寫入 elasticsearch；公司一共有 34 臺 16c64g 規格的 logstash 機器，然而這樣的高配叢集在晚高峰的時候會顯得很吃力，每晚必定會報一堆 kafka 堵塞的告警；經過調研決定使用 vector 替換 logstash，最終只用了 10 臺 16c16g 的機器便完成替換，並且之後再也沒有報過 kafka 堆積告警！