Giter Site home page Giter Site logo

backup-pruner's Introduction

Backup pruner

Backup pruner is a tool helping with regular cleanup of old backups.

It periodically scans a given directory (filesystem, S3 and SFTP are currently supported) containing the backups, removing files which are likely the least valuable ones, preventing the directory from unlimited growth.

The number of backups to keep in a given time period can be configured: you can for example define that you wish to keep 1 backup for the last day, 2 backups for the last 3 days, 5 backups for the last 10 days, 10 backups for the last month, 12 backups for the last 3 months and 15 backups for the last year. Backup pruner does not care whether the backups are created regularly or whether there are gaps and/or extra backup files. Nor does it care whether there are at least as many files as specified (in this case it will simply not delete anything in the given time period). It will do its best to keep the desired numbers of files for the given time period.

You don't need to specify the desired counts in detail. It's fine to define that for example for the last day there should be 1 backup and for the last year there should be 15 backups kept. Backup pruner will derive the optimal count for any time interval from that. It is, however, recommended to define a handful of points close to zero (e.g. a day ago, two days ago, five days ago), because even though the derived optimal counts can be a mathematically elegant function, it may behave counter-intuitively close to zero.

Using Backup pruner

Backup pruner is distributed as a Docker image. The description below assumes you know how to run Docker containers.

This is an example configuration

docker run \
    --privileged \
    --name backup-pruner \
    --rm \
    --env OPTIMAL_COUNT_CONFIG='1d: 1, 3d: 2, 10d: 5, 30d: 10, 90d: 12, 365d: 15' \
    --env STORAGE_TYPE='S3' \
    --env S3_ACCESS_KEY='AAAABBBBCCCC1111DDDD' \
    --env S3_SECRET_KEY='aaAAbbBBccCCddDDeeEEffFFggGGhhHH11112222' \
    --env S3_BUCKET='my-backup-bucket' \
    --env SCAN_INTERVAL='30m' \
    --env SEND_EMAIL='true' \
    --env LONGEST_DELAY_BETWEEN_EMAILS='12h' \
    --env SMTP_HOSTNAME='mail.example.com' \
    --env SMTP_PORT='465' \
    --env SMTP_DOMAIN='example.com' \
    --env SMTP_USER='[email protected]' \
    --env SMTP_PASSWORD='3ecre1' \
    --env SMTP_ENABLE_SSL='true' \
    --env EMAIL_FROM='[email protected]' \
    --env EMAIL_TO='[email protected]' \
    --env EMAIL_SUBJECT='Backup pruner results' \
    --env DO_PRUNE=true \
    any3w/backup-pruner

Here we define a container which will every 30 minutes (variable SCAN_INTERVAL) scan a directory on an S3 bucket (variable STORAGE_TYPE, variables S3_*). The desired counts of backup files to keep are specified the same as in one of the previous paragraphs (variable OPTIMAL_COUNT_CONFIG). The container is configured to send emails informing about the backups (variable SEND_EMAIL) using the given SMTP server (variables SMTP_* and EMAIL_*). An email will be sent if at least one file is deleted, but also if the last email was sent a long time ago (variable LONGEST_DELAY_BETWEEN_EMAILS), even if no file was deleted. The last variable (DO_PRUNE) must be present and set to "true" if any files shall be deleted. Otherwise it will do just a dry-run (determine which files to remove, send an email about it, but not actually delete anything).

The privileged mode is needed if the backup files are on S3 (needed to mount the diretory using s3fs).

This is an example of a similar configuration for a local filesystem directory:

docker run \
    --name backup-pruner \
    --rm \
    --env STORAGE_TYPE='filesystem' \
    --env OPTIMAL_COUNT_CONFIG='1d: 1, 3d: 2, 10d: 5, 30d: 10, 90d: 12, 365d: 15' \
    --env DIRECTORY='/tmp/test' \
    --env SCAN_INTERVAL='30m' \
    --env SEND_EMAIL='true' \
    --env LONGEST_DELAY_BETWEEN_EMAILS='12h' \
    --env SMTP_HOSTNAME='mail.example.com' \
    --env SMTP_PORT='465' \
    --env SMTP_DOMAIN='example.com' \
    --env SMTP_USER='[email protected]' \
    --env SMTP_PASSWORD='3ecre1' \
    --env SMTP_ENABLE_SSL='true' \
    --env EMAIL_FROM='[email protected]' \
    --env EMAIL_TO='[email protected]' \
    --env EMAIL_SUBJECT='Backup pruner results' \
    --env DO_PRUNE=true \
    -v c:/tmp/testfiles:/tmp/test \
    any3w/backup-pruner

Details of optimal count config

The optimal count config (environment variable OPTIMAL_COUNT_CONFIG) is a mandatory parameter specifying how many files should be ideally kept for a given time period.

It consists of a list of points, each point consists of a time ago specification and a count specification. The meaning of these two specifications is: for the time period between now and time ago there should be ideally count files kept.

The individual points are comma-separated (with optional white space between). Each point consists of a time ago specification followed by a colon, an optional white space and a count specification.

The time ago specification consists of a value and a unit. The supported units are: s (second), m (minute), h (hour), d (day), w (week), M (month), y (year). They are case-sensitive - pay attention to the difference between month (M) and minute (m). The units can be combined, e.g. 1h30m for one hour and 30 minutes.

Algorithm

Backup pruner parses the optimal count definition. It then plots the points in the definition into a graph, x-axis being the time ago (current time at the left) and y-axis being the count. It then connects all the points with a smooth curve. The curve represents an optimal count function (optimalCount).

Optimal count function

Backup pruner always keeps the newest and the oldest backup in the specified directory. It calculates the optimal number of intermediate files to keep as
optimalCount(oldest) - optimalCount(newest).

It then goes through all the intermediate files and for each file it calculates its value. A value of an n-th file (where n = 0 for the newest file, 1 for the second newest, etc.) if calculated as
optimalCount(n + 1) - optimalCount(n - 1). It then removes the file with the lowest value and repeats the process until the desired number of intermediate files is left.

Notification emails

If the SEND_EMAIL variable is set to true, Backup pruner will send a notification email

  • when at least one file was deleted (if DO_PRUNE is false, the email is sent if some files would be deleted if DO_PRUNE was true), and also
  • when no file was (would be) deleted, but the last notification email was sent long ago.

What it means "long ago" is defined by the LONGEST_DELAY_BETWEEN_EMAILS variable. This variable value consists of a number and a unit. The supported units are: s (second), m (minute), h (hour), d (day), w (week), M (month), y (year). They are case-sensitive - pay attention to the difference between month (M) and minute (m). The units can be combined, e.g. 1h30m for one hour and 30 minutes.

SMTP config

The following variables control SMTP:

  • SMTP_HOSTNAME - hostname, usually in the form mail.example.com or smtp.example.com
  • SMTP_PORT - port, usually 465 if SSL is enabled, 25 if not
  • SMTP_DOMAIN - domain, usually the same as hostname, but without the first bit, e.g. example.com
  • SMTP_USER - username, often identical with the sender's email address
  • SMTP_PASSWORD - password, not encrypted
  • SMTP_ENABLE_SSL - set to true to enable SSL

S3-specific settings

  • STORAGE_TYPE - to scan files in an S3 bucket, set to S3
  • S3_ACCESS_KEY - S3 access key
  • S3_SECRET_KEY - S3 secret key
  • S3_BUCKET - S3 bucket ID
  • S3_ENDPOINT - URL of the S3 endpoint if using an alternative S3-compatible server, example: http://minio.example.com:9000; do not define if using S3 on AWS (which is very likely the case)

SFTP-specific settings

  • STORAGE_TYPE - to scan files on an SFTP server, set to SFTP
  • SFTP_HOSTNAME - server hostname
  • SFTP_DIRECTORY - directory (on the SFTP server) to scan
  • SFTP_USER - SFTP server user name
  • SFTP_PASSWORD - SFTP server user password (if using password authentication)
  • SFTP_PRIVATE_KEY - path to the private key (if using key authentication) on the Docker instance (this option will be typically used in conjunction with a directory mapping -v)
  • SFTP_KEY_PASSPHRASE - private key passphrase; used if SFTP_PRIVATE_KEY is used and the key is passphrase-protected (if the key is not passphrase-protected, simply do not mention this environment variable)

example:

docker run \
    --privileged \
    --name backup-pruner \
    --rm \
    --env OPTIMAL_COUNT_CONFIG='1d: 1, 7d: 5, 30d: 8, 90d: 12, 365d: 15' \
    --env STORAGE_TYPE='SFTP' \
    --env SFTP_HOSTNAME='server.example.com' \
    --env SFTP_DIRECTORY='/var/backups' \
    --env SFTP_USER='test' \
    --env SFTP_PRIVATE_KEY='/tmp/id_rsa' \
    --env SFTP_KEY_PASSPHRASE='s3cre1' \
    --env SCAN_INTERVAL='5m' \
    --env SEND_EMAIL='true' \
    --env LONGEST_DELAY_BETWEEN_EMAILS='10m' \
    --env SMTP_HOSTNAME='mail.example.com' \
    --env SMTP_PORT='465' \
    --env SMTP_DOMAIN='example.com' \
    --env SMTP_USER='[email protected]' \
    --env SMTP_PASSWORD='s3cre1' \
    --env SMTP_ENABLE_SSL='true' \
    --env EMAIL_FROM='[email protected]' \
    --env EMAIL_TO='[email protected]' \
    --env EMAIL_SUBJECT='Backup prune results' \
    --env DO_PRUNE=false \
    -v "/user/test/.ssh/id_rsa:/tmp/id_rsa" \
    any3w/backup-pruner

Filesystem-specific settings

  • STORAGE_TYPE - to scan a directory on a filesystem, set to filesystem
  • DIRECTORY - full path to the directory to scan

General settings

  • DO_PRUNE - if set to true, the files will actually be deleted, otherwise just a dry-run will be done (even for dry-run the notification email will be sent, listing files which would be deleted)
  • SCAN_INTERVAL - how often to scan the directory; This variable value consists of a number and a unit. The supported units are: s (second), m (minute), h (hour), d (day), w (week), M (month), y (year). They are case-sensitive - pay attention to the difference between month (M) and minute (m). The units can be combined, e.g. 1h30m for one hour and 30 minutes.

Docker Hub repository

https://hub.docker.com/r/any3w/backup-pruner/

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    ๐Ÿ–– Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. ๐Ÿ“Š๐Ÿ“ˆ๐ŸŽ‰

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google โค๏ธ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.