AWS Canberra User Group

Catering this month is supplied by Little Phat Rolls.

Proving a Negative: How AWS Security Architecture Helped Us Verify a Database Extortion Scam
We received an email claiming our databases had been stolen and the malicious actors were demanding cryptocurrency in order to not leak sensitive data. In this talk we will walk through the AWS security infrastructure we have built that let us confidently call the bluff. We'll cover our setup, including bastion host architecture, comprehensive logging across IAM and VPC, root account controls, and network restrictions that ensured all data movement left a trace. During our investigation, we checked bastion logs, application logs, and root account activity to verify that no unauthorised access had occurred. As a bonus outcome: the Aws architecture helped us achieve IRAP certification.

We will share practical tips for small teams, including using CloudFormation for consistent deployments, implementing least-privilege IAM roles, leveraging AWS CloudShell to reduce attack vectors, and adopting operator/spotter processes for sensitive operations.

Sima Khatibi
Sima is a Software Developer at Aristotle Metadata and has been with the team for almost two years. She’s passionate about AWS infrastructure and enjoys tackling new technical challenges.

Harry Silke
Harry is a software developer at Aristotle Metadata and has been with the team for the past 2 years. He recently graduated from ANU with degrees in Science and Arts. He loves trail running, camping and anything outdoors!

AWS Data Centre Networking
An overview of the network architecture inside a modern data centre.
Is "north-south" vs "east-west" some kind of Feng shui? Why is the
Ethernet specification now more than two thousand pages? What are EFA and Nitro power-ups for EC2 instances?

More seriously, how should we change the way we design and build
systems in the cloud? AWS datacentres are different to the Internet
and LANs we are usually taught about and some of our assumptions and
best practices can be reconsidered. If tail latency matters for your
web applications or low latency for your machine learning system, what
AWS capabilities can help? How much can we depend on the AWS
infrastructure for redundancy and failover? I can't provide definitive
answers, but can give you ideas to think about.

Hugh Fisher
Hugh (or Hugo) Fisher is a long time network administrator
and developer of networked applications. (Don't ask him about USENET
or MBONE unless you have a lot of time to spare.) He reads Amazon
whitepapers and watches re:Invent videos in an attempt to keep up to
date with what's happening with the modern Internet.