How aakash12312/mamscn works
This repository implements SNMP polling and streaming anomaly detection, with separate scripts for training detection and classification models. The main runtime path reads device inventory, polls SNMP metrics, publishes records to Kafka, then scores and classifies them before saving results to MySQL and CSV. Training scripts consume datasets and save model artifacts. The README is empty, so this map relies on the sampled source; simulator files and unsampled integrations are not linked where usage is unverified.
Main entry points: snmp_producer.py SNMP poller · consumer_writer_snmp.py Scoring consumer · isolation_forest_train.py Isolation training · classification.py Classifier training
Components 15
Metric collection
- Device inventory
inventory.csv - SNMP poller
snmp_producer.py - Poll log
snmp_polled_data.csv - Kafka topic
Anomaly analysis
- Scoring consumer
consumer_writer_snmp.py - Model artifacts
models
Model training
- Training dataset
datasets/all_data (3).csv - Isolation training
isolation_forest_train.py - Classifier training
classification.py - Random forest training
train_anomaly.py - Legacy model
random_forest.pkl
Results and storage
- MySQL metrics
- Scored CSV
snmp_consumed_ai.csv
Other components
- Operator
- SNMP devices
Connections 17
- Operator → SNMP poller: starts
- SNMP poller → Device inventory: reads
snmp_producer.py - SNMP poller → SNMP devices: polls
snmp_producer.py - SNMP poller → Poll log: writes
snmp_producer.py - SNMP poller → Kafka topic: publishes
snmp_producer.py - Kafka topic → Scoring consumer: delivers records
consumer_writer_snmp.py - Scoring consumer → Model artifacts: loads models
consumer_writer_snmp.py - Scoring consumer → MySQL metrics: writes metrics
consumer_writer_snmp.py - Scoring consumer → Scored CSV: writes scores
consumer_writer_snmp.py - Operator → Isolation training: starts
- Operator → Classifier training: starts
- Isolation training → Training dataset: reads
isolation_forest_train.py - Classifier training → Training dataset: reads
classification.py - Isolation training → Model artifacts: saves models
isolation_forest_train.py - Classifier training → Model artifacts: saves models
classification.py - Operator → Random forest training: starts
- Random forest training → Legacy model: saves model
train_anomaly.py