jaegertracing/jaeger

Add Cassandra in healthcheck

Open

#633 建立於 2018年1月4日

在 GitHub 查看
 (10 留言) (5 反應) (0 負責人)Go (18,974 star) (2,326 fork)batch import
enhancementhelp wanted

描述

We are observing a strange issue where both Query and Collector does not reconnect to the Cassandra cluster and both logs "gocql: no hosts available in the pool"

The problem seems to be either with GoCQL or with our setup, but I think that Query and Collector healthcheck should reflect that as a problem.

As of now, killing the pods were enough to fix the issue.

貢獻者指南

技術棧
go
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。
3
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。
half day
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
Go programmingJaeger architectureCassandra basics
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。
40
研究方向
The issue proposes adding a healthcheck endpoint for Cassandra connectivity in Jaeger's Query and Collector components. Relevant files might include healthcheck handlers in `cmd/query/` and `cmd/collector/` directories. The issue comments may provide additional context on the observed failure and expected behavior. A maintainer could clarify the preferred approach for monitoring Cassandra connections.