Bintray Backuper

A tool to backup any type of public/private repositories from Bintray.

Table of Contents

• Reasons to back up repositories from Bintray
• Key Features
• Usage
• Releases

Reasons to Back up Repositories from Bintray

Backups, aka "better safe than sorry"!

• To actually own your data
• To protect from Bintray downtimes
• To prevent data loss in case Bintray loses it
• To prepare for migration to another artifact hosting service

Key Features

• Parallel downloads
• Parallel metadata resolution
• Download is skipped if checksum of local file matches expected one from Bintray metadata
• Checksum is verified for freshly downloaded file to prevent data corruption
• Timeouts, threads and buffers can be configured for better performance

Usage

bintray-backuper is released in form of a runnable .jar file, it requires JVM 1.8+ to run:

java -jar bintray-backuper.jar --subject myorg --download-dir backup-dir

Required Arguments

• --subject
  • Bintray "subject", either org or user name that hosts files.
  • Example: --subject my-org
• --download-dir
  • Directory to download files to. Files will be stored in following layout: 'download-dir/subject/repo/path-to-file'. Directory layout will be created if it doesn't exist. Checksum of existing files will be verified against metadata downloaded from Bintray.
  • Example: --download-dir backup-dir

Optional Environment Variables

• BINTRAY_BACKUPER_API_CREDENTIALS
  • Bintray API credentials in form of user:apikey string
  • To obtain API key go to "Edit Your Profile" on Bintray → "API Key"
  • Default: "" (no value, requests are made anonymously and are subject to Bintray anti-DDoS and other checks)
  • Example: BINTRAY_BACKUPER_API_CREDENTIALS="myuser:myapikey" java -jar bintray-backuper.jar

Optional Arguments

• --http-connection-timeout
  • HTTP connection timeout (seconds)
  • Default: 30 (seconds)
  • Example: --http-connection-timeout 10
• --http-write-timeout
  • HTTP write timeout for individual socket write (seconds)
  • Default: 60 (seconds)
  • Example: --http-write-timeout 10
• --http-read-timeout
  • HTTP read timeout for individual socket read (seconds)
  • Default: 60 (seconds)
  • Example: --http-read-timeout 10
• --http-call-timeout
  • HTTP call timeout (seconds)
  • Default: 300 (seconds)
  • Example: --http-call-timeout 600
• --network-buffer-bytes
  • Network stream buffer
  • Default: 16384 (bytes)
  • Example: --network-buffer-bytes 32768
• --http-threads
  • Number of threads for HTTP requests
  • Default: 6 (threads)
  • Example: --http-threads 8
• --checksum-threads
  • Number of threads for checksum verification
  • Default: number of cores * 6 (disk bound)
  • Example: --checksum-threads 32
• --checksum-buffer-bytes
  • Checksum disk stream buffer
  • Default: 16384 (bytes)
  • Example: --checksum-buffer-bytes 65536
• --download-retries
  • Number of retries to attempt for each download
  • Default: 3 (attempts)
  • Example: --download-retries 5
• --api-endpoint
  • Bintray-compatible API endpoint to use
  • Default: https://api.bintray.com/
  • Example: --api-endpoint https://mycompany-bintray-proxy.com/
• --downloads-endpoint
  • Bintray-compatible downloads endpoint to use
  • Default: https://dl.bintray.com/
  • Example: --downloads-endpoint https://dl.mycompany-bintray-proxy.com/

Releases

Binaries of bintray-backuper are published to GitHub Releases as part of Continuous Integration (triggered by git tags).

Each release contains:

• bintray-backuper.jar: executable jar
• bintray-backuper.jar.sha256: sha256 of the jar
• bintray-backuper.jar.sha512: sha512 of the jar
• dependencies.txt dependencies used to build the jar
• dependencies.txt.sha256: sha256 of dependencies.txt
• dependencies.txt.sha512: sha512 of dependencies.txt