Downloading and Uploading Backups
The Liferay Cloud backup service creates backups of an environment’s database and the full contents of the Liferay instance’s LIFERAY_HOME/data
folder. This content is stored as archive files (.gz
and .tgz
respectively) and can be downloaded via the Liferay Cloud console.
Users can also download or upload environment backups using the Liferay Cloud Console, or through Backup APIs.
The Backups page is only available in production environments for backup service versions older than 4.3.5.
Downloading Backups via the CLI Tool
Requires CLI tool version 3.12.0+, backup service version 5.9.0+, and Liferay service version 5.3.0+
You can use the CLI tool to download backups directly, including both the database and document library volume.
Run this command using a specific backup’s ID to download its database and document library simultaneously:
lcp backup download --backupId [ID]
There are three optional command line switches:
--database
specifies downloading only the database.
--doclib
specifies downloading only the document library.
--concurrency
defines the number of files to download in parallel (up to 10,000).
Here’s an example of using multiple switches:
lcp backup download --backupId [ID] --doclib --concurrency 500
The optimal number of files to download concurrently varies depending on your system and network specifications. The --concurrency
flag only affects downloads for the document library because the database is downloaded as a single file.
Uploading Backups via the CLI Tool
Requires CLI tool version 3.12.0+, backup service version 5.9.0+, and Liferay service version 5.3.0+
Database table and column names are case sensitive in Liferay Cloud. Ensure that the table names are in Pascal case before uploading your database. See Ensuring Correct Table Capitalization for more information.
Run this command to upload a backup with both a database and document library:
lcp backup upload --project [Project name] --environment [Environment] --doclib [Path to document library folder] --database [Path to database .gz file]
You must create and compress a database dump to a .gz
archive to upload it.
You can also add the --concurrency
flag with a number of streams to upload in parallel (up to 10,000).
Preparing the Database and Document Library for Upload
To upload a backup of your environment to Liferay Cloud, you must have the database prepared in its own archive file. For backup service versions before 5.9.0, you must also compress the document library in a separate archive file.
Creating the Database File
To create a MySQL dump (as a .sql
script) and compress it into a .gz
archive, run these commands:
mysqldump -uroot -ppassword --add-drop-database --databases lportal | gzip -c | cat > database.gz
If your Backup service is not updated to at least version 4.2
, you must also run the following command to convert the archive to a .tgz
file: tar zcvf database.tgz database.gz
. Then use the resulting .tgz
archive to upload.
The --databases
and --add-drop-database
flags are necessary for backup restoration to work correctly. You can also use the /backup/download
API to see how the backup service creates its MySQL dump file.
With these options, the resulting dump file contains the following code just before the create table statements.
--
-- Current Database: `lportal`
--
/*!40000 DROP DATABASE IF EXISTS `lportal`*/;
CREATE DATABASE /*!32312 IF NOT EXISTS*/ `lportal` /*!40100 DEFAULT CHARACTER SET utf8 */;
USE `lportal`;
Creating the Volume File
If you are uploading a backup with the LCP tool, you do not need to create a volume file because the tool automatically compresses your document library’s files.
To upload a backup via the Cloud console UI or API, you must compress the document library to upload it.
If permissions are not already configured for Liferay Cloud when you upload a backup, then restoring the backup to your environments afterward can take longer to complete. To avoid long restore times, navigate to your LIFERAY_HOME
folder and run this command before compressing the document library: chown -R 1000:1000 data/document_library/
.
Run this command to compress the data volume:
cd $LIFERAY_HOME/data && tar -czvf volume.tgz document_library
Downloading Backups via the Console
You can only download backups via the console before backup service 5.9.0. For versions 5.9.0+, use the CLI tool instead.
Follow these steps (as an administrator) to download a backup from the Backups page in your chosen environment:
-
Click Actions ( ⋮ ) for the backup you want to download.
-
Click on Download.
-
Click the Database (
.gz
) or Liferay (.tgz
) file to start downloading. Together, these zip archives comprise the environment backup.noteIf your Backup service is not yet updated to version
4.2
or above, then the Database volume is downloaded as a.tgz
archive instead of.gz
.
Uploading Backups via the Console
Before you can upload a backup to Liferay Cloud, you must compress the database dump and document library in separate archives. See Preparing the Database and Document Library for Upload for more information on preparing them for an on-premises environment.
Database table and column names are case sensitive in Liferay Cloud. Ensure that the table names are in Pascal case before uploading your database. See Ensuring Correct Table Capitalization for more information.
Follow these steps from the Backups page:
-
Click Upload Backup… near the top of the screen.
-
On the Upload Backup page, expand the appropriate environment, and then click the
+
icons for both the database and document library to upload them. -
When both the database dump and document library are uploaded, click Initiate Upload.
Liferay Cloud begins using the files you uploaded to generate a backup and adds it to the list you can restore to your environments. While the backup is being generated, other backups cannot be generated or restored.
A success message appears on the page when the backup is generated, and the service resumes normal operation.
Backup Service APIs
The backup service has APIs that you can also use to download and upload backups before backup version 5.9.0. You can invoke these APIs using a command line tool such as curl
.
Getting the Host Name
To invoke the backup APIs, you need the backup service’s host name. You can find this on the Services page.
The backup service’s host name is a combination of the service, project, and environment names.
Consider this example:
- Service name:
backup
- Project name:
lfrjoebloggs
- Environment name:
prd
- Host name:
backup-lfrjoebloggs-prd.lfr.cloud
Authentication
You can authenticate your request with basic authentication or a user access token.
Note that token authentication is required if SSO is enabled. You can retrieve this token from the cookie access_token
and use it with the dxpcloud-authorization
header.
Here’s an example that uses token authentication with the upload API:
curl -X POST \
https://backup-[Project name]-[Environment].lfr.cloud/backup/upload \
-H 'Content-Type: multipart/form-data' \
-H 'dxpcloud-authorization: Bearer [User token]' \
-F 'database=@/my-folder/database.gz' \
-F 'volume=@/my-folder/volume.tgz'
Passing the user token in the header dxpcloud-authorization
only works for versions 3.2.0
or greater of the backup service. Previous versions should be upgraded to at least 3.2.0
. Requests to earlier versions must use the header Authorization: Bearer [Project master token]
. You can find the project master token by running the command env | grep LCP_PROJECT_MASTER_TOKEN
in any shell in the Liferay Cloud console.
Download Database API
The API for downloading a database contains an endpoint that returns a .gz
file. The id
parameter represents the backup ID, which you can find on the Backups page. This ID is comprised of three strings separated by two dashes (e.g., dxpcloud-lqgqnewltbexuewymq-201910031723
).
Parameters
Name | Type | Required |
---|---|---|
id | String | Yes |
curl Example
curl -X GET \
https://backup-[Project name]-[Environment].lfr.cloud/backup/download/database/[ID] \
-u user@domain.com:password \
--output database.gz
If your Backup service is not yet updated to version 4.2
or above, then the Database volume is downloaded as a .tgz
archive instead of .gz
.
Download Data Volume API
The API for downloading a data volume contains an endpoint that returns a .tgz
file. The id
parameter represents the backup ID, which you can find on the Backups page. This ID is comprised of three strings separated by two dashes (e.g., dxpcloud-lqgqnewltbexuewymq-201910031723
).
Parameters
Name | Type | Required |
---|---|---|
id | String | Yes |
curl Example
curl -X GET \
https://backup-[Project name]-[Environment].lfr.cloud/backup/download/volume/[ID] \
-u user@domain.com:password \
--output volume.tgz
Upload Backup API
Follow these steps to upload a backup to Liferay Cloud with the upload backup API:
-
Invoke the backup API with the database and volume files.
Before you can use the upload API, you must compress the database dump and document library in separate archives. See Preparing the Database and Document Library for Upload for more information on preparing for the upload for an on-premises environment.
Invoking the Backup API
Parameters
Name | Type | Required |
---|---|---|
database | File | Yes |
volume | File | Yes |
curl Example
curl -X POST \
https://backup-[Project name]-[Environment].lfr.cloud/backup/upload \
-H 'Content-Type: multipart/form-data' \
-F 'database=@/my-folder/database.gz' \
-F 'volume=@/my-folder/volume.tgz' \
-u user@domain.com:password