View Code? Open in Web Editor
NEW
DyCause is a root cause analysis method for the microservice system failures.
License: Apache License 2.0
Python 5.29%
Jupyter Notebook 94.71%
dycause_rca's Introduction
🔭 I’m currently working on microservice root cause analysis, causal discovery, causal inference and multivariate time series prediction.
dycause_rca's People
Contributors
Watchers
dycause_rca's Issues
Thank you for making you code publicly available.
When running microcause.ipynb and dycause_exp.ipynb, I notice that you are using only one metric to perform the root cause analysis. Why have you not included more metrics?
Don't need real random walk, can get the rank result.
Just solve the equation.
Thank you for your implenmentation of dycause algorithm.
I wondering where files in ./data folder coming from, are they publicly available?
首先非常感谢您的辛苦工作!您复现的代码逻辑很清晰。
想请问您一下示例数据表示的是每个service在某个time slot被调用时的latency吗?
以及最后一行的view时什么意思呢?
非常期待您的回复,再次感谢您!
很棒的工作!不过有两个疑问:
MicroCause的实现,并没有按照文章里的partial correlation的方式计算Q
想问下您提到的DyCause是哪篇文章呢?是暂时还在投稿吗?
期待您的解答
首先非常感谢您的辛苦工作!您复现的代码逻辑很清晰。
想请问您一下microcause中frontend表示的是什么意思呢?
非常期待您的回复,再次感谢您