Data Integrity in SignUp, Spaces, and D!BS

We strive to keep our customer data as safe as possible, and while our servers are maintained at the highest standard, we cannot guarantee that server malfunctions will not happen at all. For the comfort of our customers, here are the ways in which we backup data and the plans we have in place should our servers malfunction.

We have a primary SQL Database server in a Virtual Private Cloud (VPC) private subnet that is cloned every night. In addition, the primary database server backs up each customer's database and ships it up to a private S3 bucket (separate server) at Amazon Web Services that is redundant and highly available (HA).

We also have individual database backups that are updated every 12 hours with a full database backup from the primary SQL server and updated every hour with a differential backup from the primary SQL server. These backups are primarily intended for catastrophic failure recovery.

• In the event of a complete failure of a single customer's database, we can restore from backup in under 5 minutes.
• In the event of a complete failure of the primary SQL server, we can spin up a clone (from the night before) in approximately 20 minutes.
• In the event of a complete failure of the primary SQL server, and a data corruption of the cloned image, we can spin up an alternate clone and rebuild the databases from the S3 backups in under two hours.

We keep daily archives for a week, weekly archives for at least a year.