We removed our free Sandbox April 25th.
You can read more on our blog.

Scheduling Backups


CLI Command examples on this page are always provided without the --application (shorthand -A) argument, assuming you’re running these commands in a connected folder (at creation or using the dotcloud connect command). For more details on connected folders, see Migrating to the CLI 0.9.

This will show you how to automatically backup services running on dotCloud.

What Can I Backup?

Below you’ll find a sample script that currently fits:

  • MySQL databases (stand-alone or master/slave),
  • PostgreSQL and PostGIS databases (the method is exactly the same),
  • services that store files in ~/data.

When you setup the script, you will have to specify the kind of backup that you need.


Making consistent MySQL backups can be tricky. If you are using MyISAM tables, the only way to guarantee consistent backups is to prevent all writes during the duration of the backup. To make sure that your backups will be consistent, the default backup script that we provide will do that. It means that for the duration of the whole backup, your database will be locked against writes. For big databases, it can mean many minutes during which you can’t UPDATE, INSERT, DELETE, etc. in your database. If you want to avoid that, you have to make sure that you use InnoDB, and update the backup script to use the --single-transaction option.

Login to the Service

The first step is to get a shell on the service that you want to backup.

This is as easy as dotcloud run -A <deployment_name> <service_name>.

For instance, to get a shell on your database server:

dotcloud run sql

Copy the Backup Script

Download this backup script for your service. The easiest way to do it is to run the following inside the service:

curl --output backup.sh http://docs.dotcloud.com/0.9/_downloads/backup.sh

Then, make the script executable:

chmod +x backup.sh

Choose Between FTP, SSH and S3

Currently, the script can upload your backups using FTP, SSH or S3.


If using FTP, you will need to know the hostname of your FTP server, as well as the login and password of the FTP user.


If using SSH, you will need to set up public key authentication on the SSH server storing your backups. Here is how to do it.

  1. Create an SSH key on the dotCloud service, by running “ssh-keygen”. Just hit “enter” when prompted for anything (it will use the default path to store the key, and will not use a passphrase to encrypt the key, which would be an issue for unattended backups anyway).
  2. Add the public key (located in ~/.ssh/id_rsa.pub on the dotCloud service) to ~/.ssh/authorized_keys on the backup server (in the account that you will use to store the backups).
  3. Test the SSH login (this is important, since it will also add the server’s public key to the known_hosts file; if you don’t do that, the backup script will fail) by doing “ssh backupuser@backupserver” from the dotCloud service. You should get a shell on the backup server.


The supplied backup script does not support SSH passwords. That’s why we use a public key. It’s usually considered to be a better practice anyway.


If you want to use S3, you need to setup your credentials on the database service first. Log into the service with dotcloud run sql bash, and run “s3cmd –configure”. You will be asked for your AWS credentials. When asked for a passphrase, don’t supply one. You can also enable HTTPS (it won’t hurt):

s3cmd --configure

This command will open up an interactive session. You can find the typical configuration below and adapt it to suit your needs:

Enter new values or accept defaults in brackets with Enter.
Refer to user manual for detailed description of all options.

Access key and Secret key are your identifiers for Amazon S3
Access Key: ...
Secret Key: ......

Encryption password is used to protect your files from reading
by unauthorized persons while in transfer to S3
Encryption password:
Path to GPG program:

When using secure HTTPS protocol all communication with Amazon S3
servers is protected from 3rd party eavesdropping. This method is
slower than plain HTTP and can't be used if you're behind a proxy
Use HTTPS protocol [No]: Yes

New settings:
  Access Key: ...
  Secret Key: ......
  Encryption password:
  Path to GPG program: None
  Use HTTPS protocol: True
  HTTP Proxy server name:
  HTTP Proxy server port: 0

Test access with supplied credentials? [Y/n] Y
Please wait...
Success. Your access key and secret key worked fine :-)

Now verifying that encryption works...
Not configured. Never mind.

Save settings? [y/N] y
Configuration saved to '/home/dotcloud/.s3cfg'

You should also create a bucket to hold your backups:

s3cmd mb s3://ramen-backups

You can of course use a different name for your bucket if you like. You can also use an existing bucket. The backups can be stored directly in the bucket, or in a directory or subdirectory.

S3 for larger than 5 GB files

If your dumps are larger than 5 GB, use s3multi instead of s3. It will split the backup into chunks, each named with a numerical suffix (.00, .01, .02, etc.) because s3cmd cannot handle uploads larger than 5 GB. To restore a multi-part dump, just concatenate the files (using cat).

Do a Manual Test Backup

The general syntax for the backup script is:

~/backup.sh <what> <how> <where>

The what will be mysql, pgsql, or data, depending on what you want to backup.

The how will be the backup method explained in the previous section.

The where is the server/bucket/credentials/... to be used.

From the SSH connection to the dotCloud service, run one of the following, depending of the method you want to use:

~/backup.sh mysql ftp ramenbackups:chowmein@ftp.noodles.com
~/backup.sh mysql ssh ramenbackups@vps.noodles.com
~/backup.sh mysql s3 ramen-backups
~/backup.sh mysql s3multi ramen-backups

Don’t forget to:

  • replace “mysql” with “pgsql” if you’re running this from a PostgreSQL service, and with “data” if you are setting up backups for a web app container and want to save its non-volatile data;
  • replace the example servers/logins/passwords/buckets with yours!

Now check your remote server or bucket: you should see a file with a naming scheme similar to ramen-default-sql-0_2011-02-10_14:17:42_UTC.sql.gz containing a SQL backup.


Since not everyone lives in the same timezone, the backup script will use UTC (or GMT) time when generating the file name. The _UTC suffix is there to remind you of this fact.

Schedule the Backup Script with a Crontab

From the SSH connection to the dotCloud service, run the following command:

crontab - <<EOF
$[$RANDOM%60] $[$RANDOM%24] * * * ~/backup.sh mysql ftp ramenbackups:chowmein@ftp.noodles.com 2>&1 | mail -s "Backup result for $HOSTNAME" cook@noodles.com

Of course, you should adapt the parameters of the backup script, just like we did in the previous step. Also, you should replace cook@noodle.com with your email address. Each time the backup script is run (once per day), you will receive an email notification. The notification will tell you the size of the backup (watch out for very small sizes, which could indicate a failure). If an error occurs, the email notification should mention it.


Using the previous recipe, your backups will run once per day. To avoid everyone starting their backups at the same time, the recipe uses a little trick: the exact time of the backup will be randomly chosen when you create the crontab. Your backup will always be started at this same time, until you modify the crontab, of course.

If you want to check which time was scheduled for your backups, just do “crontab -l”. The minute and hour of the backups are the two numbers in front of the three asterisks.

If you want your backups to occur at a fixed time, or at different frequencies, feel free to adapt the crontab.


The default emails sent by the cron daemon will contain the full command line executed. Including the password specified on the command line. To avoid disclosing your passwords, we specify MAILTO=”” so cron will never send an email by itself; and then we pipe the backup script through the “mail” command to specify the email subject ourselves.

Restoring Backups

After you’ve completed all the steps in this Guide, you should practice restoring the backup to another copy of your application. For an example with MySQL, please see the MySQL Import tutorial.