昨天讀了一篇關於未來比DevOps工程師還要更進一步走在前頭的有關SRE( Site Reliability Engineer)工程師十大特點的文章,在這邊摘錄一些重點分享一下。

  1. The SRE model has a more specific focus on application reliability and performance at scale.
  •  SRE工程師主要著力點在於應用程式在擴展上的可靠性與性能表現,該性能是一種高User Experience的指標,通常會反應在business上顧客的滿意程度。
  1. The SRE has no exclusive alignment with either Dev or Ops, and must be prepared to support either group – and stand up to them when necessary.
  • SRE雖然多數時間會與Dev一起co-work,但其實並沒有一定,他必須在有需要時跟DevOps的人一起協同工作,同時也是DevOps之間的銜接橋樑。
  1. The SRE’s principal(主要) focus is on the reliability and performance of applications – combined with a wider perspective on the needs and priorities of the business and its customers.
  • SRE需要對他的應用程式的客戶需求以及應用business有廣泛的理解,並能識別其中的優先順序,藉此引導出實作應用程式的可靠度與性能的最佳方案。
  1. Be comfortable with coding.
  • SREDev一樣熱愛寫程式並總是喜歡探求應用程式的operation之組成元件與基礎架構,如:網路與DB…等,如此才能針對應用程式的reliability performance進行改善精進。
  1. As a bridge between Development and Operations, SREs are empowered(使能夠) to fix application issues in production.
  • SRE必須能夠迅速處理production issue,並且是要能溝通DevOps一起解決問題。
  1. Error Budget - Creates an error budget of 0.01%, within which outages(斷電;中斷供應) and downtime are considered acceptable.
  • SRE應該要知道百分之百的應用程式availability reliability 是不可能的,必須有0.01Error budget容忍彈性。
  1. Today, modern applications must be able to scale quickly and have the right capacity. The SRE will keep one eye on the future, to plan for how the application can maintain performance and reliability without linear increase in manual effort.
  • SRE清楚應用程式的擴充將愈來愈快,所以必須隨時了解如何迅速擴充應用程式並且給予適當的資源容量同時一切擴充動作必須都是減少人工介入的。
  1. The SRE must have a strong drive to be a detective and understand why things are working or not working as they should. go beyond understanding everything about the application to being curious about the potential of next-gen architectures.
  • 總是了解應用程式運作的所有事不管是成功的或失敗的,並且總是能跳出應用程式框架去追求下一代新應用程式架構的樣子。
  1. They will instead proactively diagnose the problem using their holistic(整體的) knowledge-set and then get busy with coding a permanent fix, rewriting a process or working with third parties to ensure that lessons are learnt and the problem never recurs again.
  • 主動診斷並解決問題透過編程或第三方工具以及流程修正,以確保問題不會再發生。
  1. All issues are recorded in written form, stored digitally, for future reference.
  • 了解並紀錄追蹤每一個問題,從錯誤中學習。
  1. Embrace automation.
  • SRE 熱愛automation工作,把繁鎖的作業最簡化,提高工作效率。
  1. The SRE must also be able to have business centric conversations with application owners and Line-Of-Business managers, using terminology they understand.
  • SRE 必須了解Business的語言並且能轉換成技術語言來作為與business managerDevOps溝通的中繼。
  1. SREs seek to understand why and how customers interact with the business, always looking for ways to remove friction(阻力;衝突), using innovative applications and technology.
  • 了解並知道使用者如何運用business,然後從中找出應用程式不合理之處,運用技術與創新的功能解決客戶需求。
  1. Be prepared to move on.
  • 永遠不滿足現狀,在一切都運作順暢之後,知道下一步改往哪裡前進,不斷自我追求卓越。

  結論:SRE 感覺起來跟我現在工作有一點像,但是卻不是QA也不是純Operation人員,他算是對business與應用程式整體關係掌握精確的一種新角色,透過對businessapplication的知識整合來服務顧客的一種角色。

 

2018327日星期二

arrow
arrow

    jackterrylau 發表在 痞客邦 留言(0) 人氣()