/
2021-07-23 AI/ML for NFV Meeting Minutes

Anuket Project

2021-07-23 AI/ML for NFV Meeting Minutes

Jul 30, 2021

Attendees

@Sridhar Rao

@Rohit Singh Rathaur

@Girish

@Jahanvi Ojha 

@Kanak Raj

Akanksha 

@Ildiko



Sl. No.

Topic

Presenter

Notes

Sl. No.

Topic

Presenter

Notes

0

Introduction of Volunteers



Jahanvi:  IGDTU-Women . In 3rd Yr. Computer Science.

Kanak: BIT Mesra. 3rd Yr Math/Computing. 

Akanksha: BIT Mesra. 3rd Yr Math/Computing.

1

Thoth as formal project in Anuket

@Sridhar Rao

Action Item: Sridhar to send a formal mail to TSC, requesting for project approval.

From next week, Thoth will be a formal project in Anuket.

Participation in Lakelse release will be difficult. Targeting 'M' Release.

2

EUAG Update

@Sridhar Rao

We presented at EUAG meeting on 20th July. 

Action Item: To Create a page in lf-euag confluence page.

Target Completion Date: 30th July

. 

3

Review of Model (FP) enhancement Ideas

@Sridhar Rao

No progress yet.

Girish: We should continue to enhance the model, and not necessarily wait for the data. 

4

Data Status



Failure Type           

Data Model Status             

Availability           

Creation Possibility               

Node

IN PROGRESS

NO

Difficult

Links

IN PROGRESS

NO

OK

VM

YES

YES

Difficult (Experimental WIP)

Container

YES

NO

Difficult (Experimental WIP)

Application

IN PROGRESS

YES

Difficult

Middleware

IN PROGRESS

NO

Difficult

5

Volunteers - AlgoSelector 



Work was described on 19th July. Links are also project.

General Strategy Convert your questions to users in MCQ - Yes/No, multiple-choices, Scale. 

Web-Based application implementation

Step-1: Separate graphs for each of the category. Start with MIT Graph .. and enhance it based on other references.

Name

ML Category

Jahanvi                                  

Supervised                                        

Akanksha

Unsupervised

Kanak Raj

Reinforced

Step-2: Integration of these three

Step-3: Implementation platform/approach survey. Software architecture to use to implement (Ex: Python module 'abc' or dynamic forms, chat-bot).

Step-4: Implement the graph

Step-5: Testing/Review

7

Volunteers - ML Problem



Relatively easier problems to obtain data. 

Important: Volunteer should be interested in 'Publications'

Step1: Enhance the survey with "Gap Analysis"

Name                                

ML Problem                               



Trend and Pattern Analysis



Anomaly Detection



Correlation Analysis

Step2: Ideas to fill the gap.

Step3: Enhance the existing implementation

8

TVLV-Tool for FailureGen - Project Update



Work is progressing well, Next week, Demo of Stage-1.

Student: Shubhank Saxena

9 

Other Open problems (looking for contributors)



Project                                         

Duration                     

Summary                                                          

Model based, Multi-Source Data Extraction

1 Month

Given a data model, the tool should extract the required data from multiple sources (Prometheus, Influxdb, Filesystem, etc).

Data Anonymizer

1 Month

This can be part of the previous tool, or can be developed independently.

Given the format of the data-source, and columns to exclude,  the tools should create a copy of the data without those columns.

Synthetic Data Generator

1 Month

Simulate infrastructure metrics.

Emulate Collectd-Data.

Chaos-Tools and Data-Generation

1 Month

Given data-models, Identify which of the chaos tools can be reused to emulate the data.

This applies only to Kubernetes.

10

Jumphost Re-Install Status



WIP. Target Date: 25th July.