The creator of NoSQL database CouchDB says he is taking a job at Salesforce.com where he will work on a “quite ambitious” and “ridiculously cool” project related to the vendor’s cloud infrastructure.
”Beginning in January 2014 I’ll be starting at Salesforce.com and working closely with Pat Helland on a project that eventually will underpin huge amounts of their site infrastructure, improving performance, reliability and predictability, while reducing production costs dramatically,” Damien Katz said on his personal blog Monday. “It’s quite ambitious and I don’t know if I’m yet at liberty to talk about the project details and scope. But if we can pull it off we will change a lot more than Salesforce.”
A Salesforce.com spokesman declined to provide additional details of the project Katz referenced.
While Salesforce.com has long used Oracle’s database under the hood of its platform, and recently inked a new long-term deal with its rival, it has also made a number of moves in support of open-source alternatives, such as devoting development resources to HBase and acquiring Heroku, which uses PostgreSQL.
Salesforce.com doesn’t necessarily need Oracle anymore, in one observer’s view.
”It’s obvious that Oracle isn’t a great fit for Salesforce.com’s database architecture,” said analyst Curt Monash of Monash Research, via email. “Salesforce.com makes little use of Oracle’s relational features, and there are many more ways to engineer security and reliability than there were when Salesforce.com first adopted Oracle over a decade ago.”
Katz’s mention that he’ll be working with Salesforce.com architect Pat Helland may provide another clue toward the project’s nature. During a talk at the Ricon West conference in October, Helland discussed Keystone, “a proposed design that could unify and simplify many forms of storage,” according to the session abstract. “We plan to do this by building the storage to expect failure and recover quickly.”
While consumer-grade solid state disk technology costs a small fraction of what vendors charge for enterprise-grade SSD, consumer-grade SSD “will return an uncorrectable error about 100,000 times as often,” the abstract adds. “By designing the architecture of Keystone to store immutable data and surrounding that data with aggressive error checking in software, we can be extremely confident of detecting an error and fetching the desired data from one of the other places it has been stored.”