DeepLog

DeepLog is lightweight standalone but powerful log analysis command line tool.

Installation

install deep-log using pip:

$ pip install deep-log

Main Features

• search from target logs by keyword or filter condition which is written in python
• subscribe target log changes which match user-defined condition
• integrate with pandas,which can used to do data analysis based on the log content.
• customized ETL processing
• support multiple data type in filters and data analysis

Basic Usage

• search keyword

$ dl hello --target /tmp/ # search all lines in the files under folder /tmp which contain the word hello

• search with filters

$dl  --target /tmp --filter="'hello' not in _record " #search all lines in the files under folder /tmp which not contain the word hello

• subscribe log change with keyword

$dl hello -- target /tmp --subscribe #subscribe incoming change which contain keyword hello under /tmp folder

• data analysis

$dl hello --target /tmp/ --analyze="df.groupby(['_record']).size()" # find all lines which contain hello then groupby by line content

hello Jack\n     2
hello James\n    2
hello Jim\n      2
hello Joe\n      4
hello Rain\n     4
hello World\n    2

Documentation

the official documentation is hosted in https://deep-log.readthedocs.io/en/latest/

Further Steps

• add index engine to accelerate query efficiency
• integrate plot functionalities
• enhance analysis engine by introducing decision engine
• rule-based analysis template engine
• package management to deliver bundled ETL && Analysis
• machine learning integration
• support more data sources
• distributed log collections