Skip to content

backup.adoc

Latest commit

 

History

History
43 lines (35 loc) · 1.48 KB

backup.adoc

File metadata and controls

43 lines (35 loc) · 1.48 KB

Back-up and restore Timbuctoo

Backup

1. Sync databases with disc:

curl -X POST --data "backupPath={path-on-server}/datasets" http://localhost:{admin-port}/tasks/backup

adminport is 8081 by default.
When request is done, the databases of each data set will synced with disc and the contents will duplicated to folder databases.bak.

2. Copy the data

rsync -av /path/to/timbuctoo/data /target/folder --exclude databases

/path/to/timbuctoo/data is the root folder of Timbuctoo. The contents should look something like this:

auth            datasets     neo4j                  users.json
authorizations  logins.json  permissionConfig.json

Restore

The restore can be done per data set or all the data of Timbuctoo.

Whole Timbuctoo restore

  1. Copy the data to /path/to/timbuctoo/data.

  2. Rename all the folders databases.bak to databases. In /path/to/timbuctoo/data execute:

ls -d  */*/*/*/ | grep databases.bak | rename 's/databases.bak/databases/g'

  1. Restart Timbuctoo

Data set restore

  1. Find the dataset to restore. You need a user UUID and the name of the data set.

    1. You can retrieve the UUID of the user from the data set label.

      1. For the label u74ccc032adf8422d7ea92df96cd4783f0543db3b__dwc u74ccc032adf8422d7ea92df96cd4783f0543db3b is the user id.

      2. The data set name will be dwc.

  2. Paste the data of the backup to the right folder: /path/to/timbuctoo/data/{uuid-of-the-user}/{data-set-name}.

  3. Restart Timbuctoo