Skip to content

Latest commit

 

History

History
261 lines (209 loc) · 10.7 KB

README.md

File metadata and controls

261 lines (209 loc) · 10.7 KB
Raw

Onedump

Mentioned in Awesome Go GoDoc tests codecov Go Report Card License

Onedump is a database dump and backup tool that can dump different databases to different storages with a simple configuration file or CLI commands.

Features

  • Database backup from different sources to different destinations.
  • MySQL dump with zero dependencies (with built-in mysql native dumper).
  • Supports dumpers with dependencies (mysqldump and pg_dump).
  • Loads configuration from S3 bucket.
  • Compression (use job.gzip: true to enable compression).
  • Unique filename (use job.unique: true to enable unique filename).
  • Slack notification.
  • Maintained docker image that contains all dependencies.

Supported data sources

Databases Status
MySQL ✅ Suported
PostgreSQL ✅ Suported

Supported destinations

Storage Status
Local ✅ Suported
S3 ✅ Suported
Google Drive ✅ Supported
Dropbox ✅ Supported

Installation

Onedump provides both binaries and Docker images via the release process.

Binary

onedump binaries are available at https://github.com/liweiyi88/onedump/releases. Use the latest version of the binary that is suitable for your OS. After downloading the binary and moving it to a folder that is in your $PATH environment variable (e.g., /usr/local/bin/onedump), give it executable permissions (e.g., sudo chmod +x /usr/local/bin/onedump). Then you should be able to run it:

$ onedump

Docker image

Docker images are also available in Docker Hub.

When to use the Docker image

  1. You want run onedump in Kubernetes, ECS, or any other container environment.

  2. Currently, onedump does not provide a native PostgreSQL dumper out of the box. Therefore, for PostgreSQL, you need to have pg_dump installed on the same machine. The Docker image includes pg_dump, making it easier to run onedump without additional manual installation.

  3. When the current native MySQL dumper doesn't fit your needs and you need mysqldump.

Use a specific pg_dump version

The docker image contains mysql client, postgresql15-client and postgresql16-client. By default it uses postgresql16-client. However, you can pass an environemnt PG_VERSION to swtich postgresql client version between 15 and 16. For example, docker run -e PG_VERSION=15 julianli/onedump:v1.2.0-arm64 -f config.yaml.

Although we maintain both ARM64 and AMD64 Docker images, usually what you need is the AMD64 image on your production Linux machine. For example: julianli/onedump:v1.2.0-amd64*

Prerequisites

onedump provides a native MySQL dumper, allowing you to use the onedump binary to dump your MySQL database contents without needing mysqldump be installed on your machine.

If you need more advanced MySQL dump features or for PostgreSQL dump, you may want to use mysqldump or pg_dump dumper, which you need to install the below dependencies on the machine.

MySQL: mysql-client
PostgreSQL: postgresql-client

If you use the Docker image we maintain, these tools are included by default. Alternatively, you can extend or build the Docker image yourself if you need further customization.

Run onedump

onedump has just one simple command to load a config file and dump DB contents based on the configuration. It has two ways of loading the config file.

Option 1. Load configuration file from local directory

After installing onedump, you should be able to run it as a simple cli command. For example:

$ onedump -f /path/to/config.yaml

The config.yaml contains all DB backup jobs in yaml format. For all configurable items. see configuration

Option 2. Load configuration from an S3 bucket

Moreover, instead of loading your config file from a local directory, you can also store it in an AWS S3 bucket. Run the cli command to load the config file from an S3 bucket

$ onedump -f backup-config/config.yaml --s3-bucket mybucket

In this case, you pass the --s3-bucket option to indicate onedump that it should load the configuration content from an s3 bucket called mybucket. Then onedump will treat the file path option backup-config/config.yaml as the s3 key. By default, onedump will use any AWS environment variables to interact with S3, if environment variables are not found, then it will use the credentials of the default profile in your ~/.aws/credentials file. To overwirte these default credentials, you can pass --aws-key, --aws-region and --aws-secret options.

Configuration examples

For all configurable items and instructions. see configuration

Dump a local DB to 2 local directories

jobs:
- name: local-dump
  dbdriver: mysql
  dbdsn: root@tcp(127.0.0.1)/test_local
  gzip: true
  storage:
    local:
      - path: /Users/jack/Desktop/mydb.sql
      - path: /Users/jack/Desktop/mydb2.sql

Dump a remote DB via SSH and save to a local directory and a S3 bucket

jobs:
- name: ssh-dump
  dbdriver: mysql
  dbdsn: user:password@tcp(127.0.0.1:3306)/mydb
  sshhost: mywebsite.com
  sshuser: root
  sshkey: |-
    -----BEGIN OPENSSH PRIVATE KEY-----
    b3BlbnNzaC1rZXktdjEAAAAABG5vbmUAAAAEbm9uZQAAAAAAAAABAAACFwAAAAdzc2gtcn...
    -----END OPENSSH PRIVATE KEY-----
  storage:
    local:
      - path: /Users/jack/Desktop/db.sql
    s3:
      - bucket: mys3bucket
        key: backup/mydb.sql
        region: ap-southeast-2
        access-key-id: awsaccesskey
        secret-access-key: awssecret

Multiple dumps to different storage

jobs:
- name: local-dump
  dbdriver: mysql
  dbdsn: root@tcp(127.0.0.1)/test_local
  storage:
    local:
      - path: /Users/jack/Desktop/mydb.sql
- name: ssh-dump
  dbdriver: mysql
  dbdsn: user:password@tcp(127.0.0.1:3306)/mydb
  sshhost: mywebsite.com
  sshuser: root
  sshkey: |-
    -----BEGIN OPENSSH PRIVATE KEY-----
    b3BlbnNzaC1rZXktdjEAAAAABG5vbmUAAAAEbm9uZQAAAAAAAAABAAACFwAAAAdzc2gtcn...
    -----END OPENSSH PRIVATE KEY-----
  storage:
    s3:
      - bucket: mys3bucket
        key: backup/mydb.sql
        region: ap-southeast-2
        access-key-id: awsaccesskey
        secret-access-key: awssecret

Slack notification

notifier:
    slack:
      - incomingwebhook: https://hooks.slack.com/services/A0B8A11N4N/...
jobs:
- name: local-dump
  dbdriver: mysql
  dbdsn: root@tcp(127.0.0.1)/test_local
  gzip: true
  storage:
    local:
      - path: /Users/jack/Desktop/mydb.sql

Recommendation

Loading the configuration from a local directory is handy when you have control of a machine and want to run onedump as a normal cli command. However, you are responsible to make sure the config file is stored securely in that machine or maybe you are responsible for encryption at rest yourself.

On the other hand, loading the configuration from an S3 bucket is better when security is your concern (encryption, versioning and fine-grain permission control etc) and when it is not convenient to have a persistent volume to store your config file (e.g. run it via a docker container). Instead of downloading the config file from S3 to a local directory. onedump will load the config directly to memory from S3 via AWS API.

How it works

The primary use case of onedump is to run one command with a configuration file. It dumps database from different drivers to different destinations.

Connect to the database

onedump connects to your database in two ways: direct network access or SSH.

Connect the database via network access

No matter if it is to dump from your local DB or a DB host that the machine can dial directly. You can create a job configurable item in the config file

jobs:
- name: exec-dump
  dbdriver: mysql
  dbdsn: user:password@tcp(10.10.10.1)/dbname
  # the rest of config...

dbdriver and dbdsn are the required fields to be able to connect to your database. For this case, the DB host is in a private network that has an IP address 10.10.10.1. It is possible to connect to the DB when the machine that runs onedump is in the same private network.

Connect the database via SSH

You can also connect to a remote database when SSH is enabled. Create a job configurable item in the config file

jobs:
- name: ssh-dump
  dbdriver: mysql
  dbdsn: user:password@tcp(127.0.0.1:3306)/dbname
  sshhost: mywebsite.com
  sshuser: root
  sshkey: |-
    -----BEGIN OPENSSH PRIVATE KEY-----
    b3BlbnNzaC1rZXktdjEAAAAABG5vbmUAAAAEbm9uZQAAAAAAAAABAAACFwAAAAdzc2gtcn...
    -----END OPENSSH PRIVATE KEY-----
  # the rest of config...

For this case, you need to pass three extra config options: sshhost, sshuser and sshkey to tell onedump to talk to the remote database via ssh.

Save DB to storage

It is required to config at least one storage for the dump job. For example, we want it to dump a local DB and save the contents in a local directory as well as an S3 bucket.

jobs:
- name: local-dump
  dbdriver: mysql
  dbdsn: root@tcp(127.0.0.1)/db
  storage:
    local:
      - path: /Users/jack/Desktop/db.sql
    s3:
      - bucket: mybucket
        key: db-backup/mydb.sql
        region: ap-southeast-2
        access-key-id: MYKEY...
        secret-access-key: AWSSECRET..

Setting cron job

Run onedump with cron mode by passing cron experssions.

$ onedump -f /path/to/config.yaml -c 21h

The Native MySQL dumper

The MySQL native dumper provides a user experience similar to mysqldump. However, it doesn't implement all the features of mysqldump. It is suitable for most basic use cases but has some limitations:

  1. It doesn't support MySQL Spatial Data Types. For example If you use spatial data types like GEOMETRY, POINT or POLYGON, use mysqldump as the dumper.

  2. It doesn't support all mysqldump options. Currently it supports --skip-add-drop-table and --skip-add-locks

If the native MySQL dumper doesn't meet your needs, switching to mysqldump is easy. Just update dbdriver in the configuration file to mysqldump:

jobs:
- dbdriver: mysqldump
  ...