AWS Error Exposed GoDaddy Business Secrets (zdnet.com)
Internal information belonging to hosting provider GoDaddy has been exposed via an error in Amazon's AWS bucket configuration. According to cybersecurity firm UpGuard, a set of documents were left in an Amazon S3 bucket which was available to the public. ZDNet reports: The information involved in the security breach appeared to describe GoDaddy's architecture, as well as "high-level configuration information for tens of thousands of systems and pricing options for running those systems in Amazon AWS, including the discounts offered under different scenarios," according to UpGuard. Configuration files for hostnames, operating systems, workloads, AWS regions, memory, CPU specifications, and more were included in the exposed cache, which described at least 24,000 systems.
"Essentially, this data mapped a very large scale AWS cloud infrastructure deployment, with 41 different columns on individual systems, as well as summarized and modeled data on totals, averages, and other calculated fields," the cybersecurity firm said. The open bucket, called "abbottgodaddy," also included what the company believes to be business information relating to GoDaddy and Amazon AWS' relationship, including rate negotiations. This information should have been kept confidential. The open bucket, called "abbottgodaddy," also included what the company believes to be business information relating to GoDaddy and Amazon AWS' relationship, including rate negotiations. This information should have been kept confidential.
"Essentially, this data mapped a very large scale AWS cloud infrastructure deployment, with 41 different columns on individual systems, as well as summarized and modeled data on totals, averages, and other calculated fields," the cybersecurity firm said. The open bucket, called "abbottgodaddy," also included what the company believes to be business information relating to GoDaddy and Amazon AWS' relationship, including rate negotiations. This information should have been kept confidential. The open bucket, called "abbottgodaddy," also included what the company believes to be business information relating to GoDaddy and Amazon AWS' relationship, including rate negotiations. This information should have been kept confidential.
Don't name your instances after your company. Select a neutral naming policy, and follow it.
It is bad enough when business data gets leaked, but why put a big flag on anything accidentally visible to let people know that they should snoop into it because they're not supposed to see it?
Nekkid pictures of Danica Patrick ?
errr....umm...*whooosh* *whoosh* Is this thing on ?
I'm confused should this information should have been kept confidential? I wish the summary would clarify that
It's too easy to have an epic security fail because it's too complicated, the peons configuring it don't know WTF they're doing and management doesn't care as long as it involves the words "cloud" and "savings".
I forgave the local semi techy tabloid type online newspaper getting this wrong when the news broke days ago, but this is Slashdot FFS. This was user error, not a systems error. An AWS employee fucked up.
que?
It's unambiguously an error in the AWS bucket configuration. The title doesn't state responsibility but the first line in the brief description does.
how many times does a bucket need to be exposed before people get smart enough to. stop. fucking. using. them. for. important. data.
The editors should ensure that duplicate sentences are removed from the summary. The editors should ensure that duplicate sentences are removed from the summary. It seems like this could be automated. It seems like this could be automated. I don't know why I see articles about 10% to 20% of the time with duplicate sentences, or sometimes even whole paragraphs, in them. I don't know why I see articles about 10% to 20% of the time with duplicate sentences, or sometimes even whole paragraphs, in them. It's kind of annoying. It's kind of annoying.
The article states that an AWS employee created the bucket and did not follow best practices. That's part of the problem with AWS, they apply the "professional" tag to new hires who sometimes only have 2-3 months experience with the company.
"Tom Bombadil is a merry fellow! Bright Blue his jacket is, and his boots are yellow!" -Tom Bombadil
Calling this an AWS error is like calling leaving an unlocked S series in an airport parking lot a "Mercedes Benz" error. This is an error that GoDaddy, who is typically poor when it comes to security knowledge, made in their implementation. They used S3 without applying proper policies around it, all of which are more than sufficiently documented.
We have two major errors here, or three, all by AWS and their employees.
An AWS employee created this bucket with it set to be accessible by everyone in the world.
Why did the set it to be accessible to anyone and everyone? They didn't set it, that's the DEFAULT on AWS. AWS says Security Groups is their implementation of a firewall. Just about the only firewall that defaults to wide open, no security at all.
Amazon's sales process apparently involves employees manually clicking on the generic UI to create resources for each customer, rather than having a Cloud Formation or other script for "create customer file".
There are many failures here. The failure of the AWS employee to override the really stupid defaults is, to me, the least significant. Had he never been hired, other employees would be frequently creating buckets with the default, and stupidly insecure, settings. Had the default been to be secure, and let the let customer select places it SHOULD be accessible from, this wouldn't have happened and it wouldn't be a common occurrence.
I've spent 20 years full time IT security. I'm very much security aware. I've forgotten to change the incredibly stupid AWS defaults, thereby leaving a security problem.
An audit script that runs as new resources are created, and the again periodically, is on my to-do list. We're a small team with hundreds of dollars per month of AWS spend. If we can do it, certainly AWS themselves can and should have such a script for internal use.
Yes, under their Customer Responsibility Model, AWS says the customer is responsible carefully avoiding the dangers inherent in the design of their product. That was a common, and arguably legal, position to take until 1963.
Since 1963, you may have noticed that self-propelled lawnmowers have deadman switches that automatically shit off the mower and apply a brake if the operator let's go of the handle for a second. All openings are covered by a metal plate held closed with a powerful spring any time the attachment isn't attached. In general, products have every reasonable safeguard in place in order to eliminate any foreseeable danger. That's because companies are legally required to put safeguards in place to guard against dangers whenever it's reasonably possible. "It's the customer's responsibility to avoid the clear dangers we designed into our product" hasn't been a legitimate or legal stance since at least 1963.
The title suggests Amazon screwed up, when it was GoDaddy that screwed up their usage of Amazon's services. Kinda shady. The title.
This is a perfect example why companies refuse to outsource their Exchange servers, SharePoint boxen, as well as file storage services to the cloud. HR won't let them for reasons such as this.
For those arguing it is GODADDYS FAULT not Amazon they miss the point. It was perfectly secure and fine the way it was. Some bean counter who is not trained in risk management urged to fire their IT workers and outsource this function to the cloud to save money.
They got what they paid for.
http://saveie6.com/
"Although the potential threats to exploit this kind of data require intentional malicious actors, the exposure of that data through misconfigured storage does not," UpGuard said. "From operations as large as GoDaddy and Amazon, to small and medium organizations, anyone who uses cloud technology is subject to the risk of unintentional exposure, if the operational awareness and processes aren't there to catch and fix misconfigurations when they occur."