Skip to content

Latest commit

 

History

History
90 lines (67 loc) · 4.25 KB

README.md

File metadata and controls

90 lines (67 loc) · 4.25 KB
Raw

Scanner system

This scanner app built as universal and flexible platform for scan data from online resources.

Features

  • Global and detailed config settings

    • System can be flexibly configured without changing code (zero code config)
    • Customer can configure all parts of system such as
    • global settings like a timezone, proxies list & user_agent settings and theirs rotation rules, cache settings and so on
    • log target, log's rotation rules, format etc.
    • scan rules and data normalization settings
    • storage setting, easy to change storage to different type, like SQL or NoSQL storage
    • easy to add new types of crawlers regarding universal interface
    • chose your own system wide timezone
    • monitoring settings (i'm alive reports)
    • auto-recovery after server crash settings

  • Universal Transport Layer

    • System wide support of proxies (including rotation proxies by Round Robin algorithm during scan process). Can be disabled any time by customer.
    • System wide support of substitution User Agents while running (by Round Robin algorithm). Can be disabled any time by customer.

  • Storages

    • easy to change storage
    • can support different storage types (like sqlite, postgres, redis,..)

  • Logs

    • use your own storage for logs
    • short information about successful scans and detailed debug info about failed scans
    • set up log rotation and log compression conditions to keep storage free

  • Crawlers

    • add easily own crawlers by unified crawlers interface
    • configure all crawler's settings separately if you need with config file
    • polite scheme of scanning data to prevent blocking
    • configure scan parameters like scan interval or delays between requests

  • Anti-scan protection

    • To avoid anti-scan mechanisms of websites system use connections with the proxies and rotation of User Agent identification

  • Avoid extra requests from the websites by using polite mechanisms of downloads

  • Auto-recovery after server crash

    • In case of server unexpectedly crash next start of scanner system will recover own state and system idempotency
    • No extra downloads when something went wrong

  • Flexible configured i-am-alive reports to outer service

  • Use docker or docker-compose to start system in container

    • $ docker build -q -t scan_app . && docker run scan_app
      or
    • $ docker-compose up

  • All code covered by tests

Settings

System-wide settings stored in settings.cfg file and have detailed descpiptions

Transport Layer

Add proxies lines to file list_proxy.cfg and set up option RotateProxies = yes in section 'CRAWLER.*' to uses rotating proxies mechanism. Also you can add User Agents strings to file list_user_agents.cfg and set up RotateHeaders = yes in the same section to enable UA rotating tool. In this case every connection will be established on new ip with new UA parameter.

Anti-scan protection

Use RotateProxies = yes and RotateHeaders = yes like it described in the 'Transport Layer' section to activate anti-scan protection mode

Logs

Set up parameters in the LOGS section of the settings.cfg file

Storages

System uses realisation independent storage interface so you can easily add new type of storage to system without any difficulties To add new storage just add new class inherited from StorageInterface and swith on connection in the Storage class (storages/storage.py).

Crawlers

System uses realisation independent crawlers interface and you can add new crawler easily. Add new crawlers class inherited from CrawlerInterface and plug it to Crawler class (crawlers/crawler.py) Scanner uses his own mechanism to decrease load into target's website by preventing repeated requests

Data normalization

Scanner system use DataStructure class (common.py) to unify sheduler of tasks. All datum will be normalize automatically Add your own rules in the 'RULES' section

Monitoring

Every ScanNewTasksInterval seconds system send alive status to AliveStatusURL ('CRAWLER.*' section of settings.cfg file). Option can be disabled

Auto-recovery

If something went wrong (for example k8s restart destroyed and created instance) scanner system can obtain own previous state automatically. Of course you have to use storage outside your instance