microsoftgraph / microsoftgraph/msgraph-sdk-python
Improve execution time on Model deserialization
還沒有人認領這個 Issue。
- 主要語言
- Python
- 星號
- 630
- 分支
- 96
- 平均合併
- 15 小時 20 分鐘
- 30 天內合併 PR
- 3
描述
So I tried to migrate our usage of raw API calls to Microsoft Graph to your new release 1.1.0.
And I measured execution time with the bare-bones method time.time() before method call, time.time() after method call.
I've tried performing get requests for the endpoints: me, users, domains.
The results were that the sdk is on average 3.5 times slower than the raw call to API which I think a great deal.
Furthermore, I was curios where this overhead came from and I searched through the code with debug.
And the most time spent in execution is after the call to the rest api, when the sdk is building a model by deserializing the json object got from the request in a sdk specific object.
This execution time difference lead me to decide that I should postpone the migration until this is improved or fixed.
貢獻指南
從這裡開始
- 先讀完整個 Issue，再讀專案的貢獻指南。
- 在 Issue 下留言說明你要接手 —— 這能避免兩個人做同樣的事。
- Fork 儲存庫，在一個分支上完成修改。
- 送出 Pull Request，並在描述裡引用這個 Issue 編號。
研究方向
首先追蹤 REST 回應後 SDK 的模型反序列化流程，issue 將其識別為開銷的主要來源。針對 me、users 和 domains 端點重現報告中的比較；當能夠證明相較於直接 API 呼叫，執行時間有可量測的減少時，即視為完成。
由索引模型根據 Issue 內容生成。
評估
- 技術堆疊
- python
- 領域
- api, performance
- Issue 類型
- 重構
- 難度
- 4/5
- 預估耗時
- 3-5 天
- 活躍度
- 停滯
- 描述清晰度
- 基本清楚
- 新手友好度
- 35/100