SODEW/sodew-bash-main/docs/restore.md

KUBE  /  Restore

Restore

The restore process is divided into three stages:

1. Restoring website files from the <domain> directory or the <domain>.tar.gz archive.
2. Restoring the <domain>.conf file.
3. Restoring the database from the <domain>.sql file or the <domain>.sql.tar.gz archive.

Warning

If an error occurs at any stage, the process does not stop.

Warning

Files, databases, and other resources at each stage are either pre-cleaned or immediately overwritten without confirmation.

Commands

-r, --restore <domain> [action]

Restore site files and database from backup for a domain.

Example:

sudo kube.sh -r example.com

The source of resources for recovery is the directories defined in backupPath and backupLongTermPath.
In the backup folder is searched for 3 types of resources:

• <source path>/<date>/<marker>/<domain> - directory with site files (file:d)
• <source path>/<date>/<marker>/<domain>.tar.gz - site file archive (file:a)
• <source path>/<date>/<marker>/<domain>.sql - site database SQL-file (db:f)
• <source path>/<date>/<marker>/<domain>.sql.tar.gz - site database archive (db:a)
• <source path>/<date>/<marker>/<domain>.conf - site database archive (conf)

The search results in a list of format: <counter>: <data> <marker> [<list of resource>]. Then you should specify the number of the selected marker for restore.

Note

If one resource is selected (file:d/file:a/db:f/db:a/conf), only one resource will be restored. The other will remain unchanged.

Note

If there's a file:d and a file:a. Priority is given to file:d. If there's a db:f and a db:a. Priority is given to db:f.

Note

Once database are restored, the Redis keys for the domain are deleted.

Example:

1: 2025-04-29 14-22-22 [file:d file:a db:f conf]
2: 2025-04-29 14-20-20 [file:d db:a]
3: 2025-04-29 12-00-00 [file:a conf]
4: 2025-04-29 _first [db:a]

You can specify an additional parameter after the domain:

• list - will display a list of available markers for restore
• last - automatic site restore from the last backup
• full - automatic site restore from the last FULL backup
• auto - automatic site restore from the last FULL backup or the last backup
• N - automatic site restore from the last backup #N (N: 1+)

FULL backup is a backup that contains a copy of the site files, a copy of the database, and a copy of the .conf file
last backup #N marker number (starting from 1) in the list sorted by creation time (can be seen with the list command)

Example:

sudo kube.sh -r example.com list
sudo kube.sh -r example.com last
sudo kube.sh -r example.com auto
sudo kube.sh -r example.com 3

Warning

Restoring from the last full copy will be done without question.