atlassian-support-cleaner

CLI tool to clean atlassian support.zip from various data

Description

Even if Atlassian cleans some data out when creating a Confluence support zip, there still remains plenty information (amongst others also personal information) about your system and users.

This tool is designed to clean out the base URL and Usernames (see How it works what this means exactly).

(May work with other support zip files, but is only tested for Confluence.)

Requirements

Python ≥ 3.5

Usage

./supportcleaner.py BASEURL [--supportzip SUPPORT_ZIP] [--filterfile FILTER_FILE]

Environment

MAX_TMP_DIR_SIZE: Set maximum allowed size in bytes of temporary directory for extraction. DELETE_AFTER_DAYS: Set maximum age for files, after which they should be flagged for deletion.

Example

./supportcleaner.py my-base-url.net --supportzip Confluence_support_2019-02-14-11-43-28.zip

How it works

This tool does cleaning the support zip by

• extracting the zip file or giving instructions for placing relevant files manually in the temporary directory, if no supportzip is supplied
• replacing in files (defined via LOGDIR, default: all):

Example	Replacement
subdomain.BASEURL	URL_CLEANED
userName: NAME_WITHOUT_SPACE	userName: USERNAME_CLEANED
customer@business.com	EXTERNAL_EMAIL_SHA256:b6c4586387_CLEANED
employee@URL_CLEANED	INTERNAL_EMAIL_SHA256:2fdc017705_CLEANED
*.smhss.de	SMEDIA_DOMAIN_CLEANED
127.0.0.1	127.0.0.IP_CLEANED
----- BEGIN RSA PRIVATE KEY ----- ...	PRIVATE_KEY_CLEANED
----- BEGIN CERTIFICATE ----- ...	CERTIFICATE_CLEANED
Big Business LLC	BUSINESS_CLEANED
Herr Müller	NAME_CLEANED

• remove files that are older than DELETE_AFTER_DAYS (default: 180) days (interactive)

• remove files that are within the LARGEST_PERCENT of files (default: largest 10%) (interactive)

• you can add additional filters by using the filterfile argument, which takes a filepath:

  • each line of the file is treated as one filter
  • each filter is splitted by "||" in two parts
  • first: search pattern (Python regex), second: replacement

• if supportzip was supplied: creating new zip named "cleaned.zip" (if already existing, it will be deleted at program start)

Predefined filters

Several filters are already defined in the filters.txt. These can and should be modified and extended to suit your specific needs.

As is, there are filters for cleaning:

• Usernames (in a specific format, replacement includes truncated SHA256 hash)
• Baseurl (including subdomains)
• Mail addresses (differentiates between internal (contains URL_CLEANED) and external. Replacement includes truncated SHA256 hash of the local part)
• smhss.de-Domains
• IP-Addresses (only the last byte is redacted)
• Private keys and certificates (in a specific format)
• Business names
• Personal names in addresses (starting with 'Frau' or 'Herr')

Examples for these filters can be seen above. Some extended documentation is provided with the regexes in filters.txt.

Notes

Please be aware, that due to creating a new zip all original timestamps as well as file owners and permissions are lost.

Disclaimer

Please be aware, that this tool cannot guarantee to cleanup all personal information or information related to your company.

License

MIT © 2019-2020 //SEIBERT/MEDIA GmbH