vespa-engine/vespa

Expose NEED_RESTART as a metric

Open

#4,631 建立於 2018年1月11日

在 GitHub 查看
 (0 留言) (0 反應) (0 負責人)Java (4,948 star) (561 fork)batch import
enhancementgood first issueneed help

描述

After a meeting today with Zedge we concluded that only providing information about the need to restart services upon deploy is not enough.

When config changes are deployed that require services to be restarted the services should themselves report that either in health or as a metric.

貢獻者指南

技術棧
java
領域
backendobservability
議題類型
feature
難度面向新貢獻者的預計實作難度，1 表示很小改動，5 表示專家級工作。
4
預計時間有經驗貢獻者完成調查、實作、測試並準備 pull request 的粗略時間範圍。
3-5 days
活動狀態議題目前的可參與程度：新鮮、活躍、陳舊、阻塞或等待維護者輸入。
stale
清晰度議題是否清楚說明預期改動、驗收標準和下一步。
mostly clear
前置要求
Java knowledgeVespa internalsMetrics system understanding
新手友善度1-100 的估計分數，表示該議題對首次貢獻者的友善程度。
20
研究方向
Examine Vespa's existing metric reporting infrastructure, particularly how services expose health and metrics via REST APIs. Look at the config server or node repository code to understand where NEED RESTART state is tracked. The goal is to add a metric (e.g., a gauge) that reports the restart requirement status. Consider extending the /state/v1/health endpoint or adding a custom metric to the metrics API. Review similar past contributions for patterns.