Slashdot Mirror
Amazon Will Be Off All Oracle Databases By End of 2019, Says AWS Chief
Amazon Web Services CEO Andy Jassy said in an interview on Wednesday that almost all of Amazon's databases that ran on Oracle will be on an Amazon database instead. "We're virtually done moving away from Oracle on the database side," Jassy said. "And I think by the end of 2019 or mid-2019 we'll be done." CNBC reports: Amazon is reducing its reliance on Oracle for its data needs and is instead using its own services. Jassy said 88 percent of Amazon databases that were running on Oracle will be on Amazon DynamoDB or Amazon Aurora by January. He added that 97 percent of "mission critical databases" will run on DynamoDB or Aurora by the end of the year. On Nov. 1, Amazon moved its data warehouse from Oracle to its own service, Redshift, Jassy said.
They offered an alternative to IBM that many considered to be a good choice at the time .. much better
Today, it's just expensive and old, while the competition got better
Times change
Amazon Aurora is technically Oracle-based, depending on how you look at it, since it's based off MySQL, which is currently an Oracle product. According to https://www.percona.com/blog/2..., it is based off the MySQL database source code of 5.6.10, which was released 2013-02-05, 3 years and 1 month after Oracle purchased Sun, which is 2 years after Sun bought MySQL.
I recall reading about Aurora this YEARS ago. (at the time, at least) Aurora is really just a proprietary storage engine that they dropped into MySQL.
They now have versions "compatible with" both MySQL and PostgreSQL.
The PostgreSQL one is the one they are using internally!
You're in luck! Amazon Aurora is basically a different backend for PostgreSQL (just announced) or MySQL that simplifies management. You don't have to learn anything new.
And that is important, because if they can move away from Oracle so can anybody.
Uhm, no. I mean, yes, Oracle isn't increasing their customer base, but saying "if Amazon can do it, anybody can do it" misses the three big reasons why Amazon can do it.
First, Amazon has billions of dollars at their disposal. Even if Oracle was letting them run their database for zero dollars and it was nothing more than a dick waving competition between Jeff Bezos and Larry Ellison just so Jeff could show Larry that he could, Amazon can financially afford to do that.
Second, Amazon has the coding talent to do it. A whole lot of people using Oracle are still doing so because they don't have the specialists required to do that sort of migration. Even if they did, most Oracle customers run Oracle because an upstream piece of software relies on it, so even if they wanted to retool *and* they had a sufficiently skilled DBA to move the data over, they probably don't have the ability to do the same for their upstream software. Pursuant to the prior point, Amazon can either fix it themselves (because they wrote it), demand the upstream vendor retools for DynamoDB (because they can afford it), or they can write a replacement that fits well enough to route around it.
Finally, unlike most Oracle customers, Amazon can easily recoup their expenses for writing DynamoDB - not just in the money they save by not-paying Oracle, but by selling the use of the database on AWS. That's fairly unique to Amazon; most other Oracle customers aren't selling database-aaS such that rolling their own will pay dividends.