Google Apps: Cloud Disaster Recovery, Backup Better Than On-Premise

With that ominous question, Google spotlighted the disaster recovery and data backup capabilities of its Google Apps cloud computing platform and unveiled a new disaster recovery feature to Google Apps: synchronous data replication, which is the real-time replication of data in multiple servers and data centers.

Synchronous data replication ensures every piece of data entered or changed in any Google Apps application is copied and backed up in almost real-time for disaster recovery. It makes copies of digital files, propagates them across various copies and breaks it into small pieces to be stored on servers across the globe in Google data centers.

With Google Gmail suffering occasional outages sparking massive user outrage, and other data stored in the cloud via Google Apps appearing ripe for the picking, businesses are often hesitant to hand over the keys to their data and have it stored elsewhere. And with other cloud outages grabbing headlines, VARs and users are stuck trying to decide whether in the cloud or on-premise are the way to go.

Google is trying to alleviate some of the fears surrounding moving and keeping data in the cloud. In a blog post this week Google Apps Senior Project Manager Rajen Sheth hammered home that Google offers disaster recovery and data backup to its Google Apps portfolio, which comprises Gmail, Google Docs, Google Calendar, Google Groups, Google Sites and Google Video, for no additional cost to paying Apps users. And while Google has used synchronous replication in Gmail for the past few years, Google has now extended the disaster recovery feature to include all of the applications in the Google Apps portfolio, Sheth said.

Basically, Google said that when Google Apps offerings like Gmail go down, data isn't lost and synchronous data replication helps. Google, in so many words, also said cloud-based disaster recovery, backup and replication solution trump their on-premise counterparts in three major categories: cost, risk and ease.

Sheth said that Google Apps' data recovery and backup options are stronger than on-premise solutions. For e-mail, he wrote, small businesses usually have a mail server and copy e-mail data to tape at daily or weekly intervals. When something goes wrong, the tapes are used to restore the data that was saved prior to the most recent backup. "But the information created after the most recent backup is lost forever," he said.

Some goes for large businesses that use a storage area network (SAN). Sheth said SANs are pricey and offer little assurance if the data center goes down. A disaster recovery site, he added, requires a new data center, identical mail servers, another SAN and more staffing resources. And if a disaster hits both data centers "you're toast," Sheth wrote, adding that to avoid that the second data center is usually built at a distance creating headaches and requiring massive amounts of data to move back and forth potentially clogging the network.

"There are other backup options as well, but the story's the same: as redundancy increases, coast and complexity multiplies," he said.

Sheth said Google Apps users don't have to worry about losing the data they create and store in Google Apps.

Disaster recovery is usually measured in terms of Recovery Point Objective (RPO), which represents the amount of acceptable data loss in the event of an outage or the gaps between backups, and Recovery Time Objective (RTO), which represents the acceptable amount of downtime before service is restored. Sheth said companies running SANs have RPO and RTO targets of an hour or less. Large companies can pay more to lower that number, he said, but SMB's are pretty much out of luck.

"For Google Apps customers, our RPO design target is zero, and our RTO design target is instant failover," Sheth wrote. "We do this through live or synchronous replication: every action you take in Gmail is simultaneously replicated in two data centers at once, so that if one data center fails, we nearly instantly transfer your data over to the other one that's also been reflecting your actions."

Sheth said Google's goal is to not lose any data when it's moved from one data center to the other and to transfer data quickly so any interruption goes unnoticed. He admits, however, that nothing is perfect, but Google works to offer the same data replication levels for all applications like Gmail, Calendar, Docs and Sites.

He said some companies have added on-premise synchronous data replication, but that gets costly. For example, he said, to backup 25GB of data with synchronous replication can run anywhere from $150 to $500 in storage and maintenance costs per employee plus the cost of applications.

Meanwhile, Google's new synchronous replication in Google Apps is possible because Google runs several large data centers for millions of users, which reduces costs and increases resiliency and redundancy, while loads are balanced between data centers as needed. Additionally, Google has high speed connections between its data centers for swift transfers, which allows for the replication of large amounts of data simultaneously.

Sheth wrote that with synchronous replication "Google Apps gives companies of all sizes access to technology that until recently was available only to large enterprises. And it's available at a dramatically lower cost than the on-premise alternatives, without the usual hassles of upgrading, patching and maintaining the software."