Skip to content

Latest commit

 

History

History
131 lines (87 loc) · 5.56 KB

README.md

File metadata and controls

131 lines (87 loc) · 5.56 KB
Raw

DumpDB

Codacy Badge

DumpDB imports credential dumps into a database to improve search performance.

There are two types of databases that will be created; one type stores the breach sources and the other type stores the dumped records. There should be a single sources-type database which stores where each record comes from (e.g. it could come from adobe2013 or collection1). There will be one or more databases which store the dumped records, these will be indexed and searched seperately.

Installation

This project requires Go version 1.15 or later. You will also need access to a MariaDB (recommended) or MySQL server.

go get -u github.com/darkmattermatt/dumpdb

Example Usage

Initialise the databases

go run github.com/darkmattermatt/dumpdb init -c "user:pass@tcp(127.0.0.1:3306)" -s sources -d adobe2013,collection1

Import the dumped data

go run github.com/darkmattermatt/dumpdb import -c "user:pass@tcp(127.0.0.1:3306)" -s sources -d adobe2013 -p adobe /path/to/data.tar.gz /more/data.txt
go run github.com/darkmattermatt/dumpdb import -c "user:pass@tcp(127.0.0.1:3306)" -s sources -d collection1 -p collections /path/to/data.tar.gz /more/data.txt

Search the indexed data

go run github.com/darkmattermatt/dumpdb search -c "user:pass@tcp(127.0.0.1:3306)" -s sources -d adobe2013,collection1 -Q "email LIKE '%@example.com' LIMIT 10"

General Info

Verbosity:

Output levels are as follows:

  1. FATAL: Only show errors and search results
  2. RESULT: Only show errors and search results
  3. WARNING: Nonfatal errors (usually occurring in one of the query threads)
  4. INFO: The default level, provides minimal information at each step of the process
  5. VERBOSE: Tells you what's going on
  6. DEBUG: Spews out data

Global Parameters:

  • config='': Config file
  • v=3: Verbosity. Set this flag multiple times for more verbosity
  • q=0: Quiet. This is subtracted from the verbosity

Init

Initialise a database for importing.

Parameters:

  • databases+: One or more positional arguments of databases to initialise
  • databases="": Comma separated list of databases to initialise
  • conn=: connection string for the MySQL. Like user:pass@tcp(127.0.0.1:3306)
  • sourcesDatabase="": Initialise the following database as the one to store sources in
  • engine="Aria": The database engine. Aria is recommended (requires MariaDB), MyISAM is supported for MySQL
  • indexes="email_rev": Comma separated list of columns to index in the main database. Email_rev is strongly recommended to enable searching by @email.com

Process

Process files or folders into a regularised tab-delimited text file.

Parameters:

  • filesOrFolders+: One or more positional arguments of files and/or folders to import
  • parser=: The custom line parser to use. Modify the internal/parseline package to add another line parser
  • batchSize=4e6: Number of lines per output file. 1e6 = ~64MB, 16e6 = ~1GB
  • filePrefix="[currentTime]_": Temporary processed file prefix

File Processing

  • .tar.gz, .tgz: Decompress and open tarball, process each file
  • .txt, .csv: Create bufio.Scanner
  • bufio.Scanner: Process each line

Import

Import files or folders into a database.

Parameters:

  • filesOrFolders+: One or more positional arguments of files and/or folders to import
  • parser=: The custom line parser to use. Modify the internal/parseline package to add another line parser
  • conn=: Connection string for the SQL database. Like user:pass@tcp(127.0.0.1:3306)
  • database=: Database name to import into
  • sourcesDatabase=: Database name to store sources in
  • compress=false: Pack the database into a compressed, read-only format. Requires the Aria or MyISAM database engine
  • batchSize=4e6: Number of results per temporary file (used for the LOAD FILE INTO command). 1e6 = ~64MB, 16e6 = ~1GB
  • filePrefix="[database]_": Temporary processed file prefix

Notes:

  • By default, only the mysql user is able to read/write to the database file directly. A workaround is to run go build . and then sudo -u mysql ./dumpdb import ...
  • Only files with whitelisted file extensions are processed (to avoid trying to import a binary file as a text file). Currently supported extensions are .tar.gz, .tgz, .txt, .csv.

Search

Search multiple dump databases simultaneously.

Parameters:

  • query="": The WHERE clause of a SQL query. Yes it's injected, so try not to break your own database
  • columns="all": Comma separated list of columns to retrieve from the database
  • conn=: Connection string to connect to MySQL databases. Like user:pass@tcp(127.0.0.1:3306)
  • databases=: Comma separated list of databases to search
  • sourcesDatabase="": Database name to resolve sourceIDs to their names from

Notes:

  • The query is injected into the SQL command which means that any LIMIT statements are applied per database

External Libraries

This project makes use of several excellent open-source libraries, listed below: