Cloud architecture past the happy path.
Mobycast turns AWS and distributed-systems topics into patient, practical architecture sessions. John Christensen and Chris Hickman work through services like S3, CloudFront, DynamoDB, ECS Fargate, and VPC networking with an engineer's attention to edge cases.
PRESS PLAY
Mobycast is a long-running technical show about the machinery behind cloud-native software. Its center of gravity is AWS, but the episodes are less about service announcements than about how systems are actually put together. A simple image-hosting service becomes a two-part architecture exercise. Uploads go one way. Downloads become a different problem. S3, CloudFront, DNS aliases, custom domains, TLS, transfer acceleration, and global edge locations all get treated as parts of one design rather than isolated features. The show is strongest when it slows down around the friction points that tutorials often skip. DynamoDB episodes move from NoSQL data modeling into partitioning, replication, leader election, Paxos, and the limits of consistency. Those conversations also connect modern managed services to older internet-scale database ideas through Chris Hickman's story of Leviathan. VPC episodes bring the same patience to private subnets, VPN access, NAT, TCP and UDP headers, MAC addresses, and the hidden work of routing traffic safely. ECS Fargate episodes test the appeal of serverless containers against the operational details of IAM roles, launch types, scaling, and trade-offs. John Christensen often plays the role of the curious builder, stopping on points that feel obvious only after someone explains them carefully. Chris Hickman supplies the systems-level explanation and keeps the architecture honest. The pacing is conversational, with occasional family-life detours and light banter, but the technical work remains the point. Mobycast does not flatten cloud architecture into slogans. It follows the chain of consequences. That makes it useful for engineers who already know the names of AWS services but want a better mental model for when, why, and how to use them. The show rewards listeners who like diagrams, failure modes, and implementation nuance. It is practical without becoming shallow. It is detailed without feeling like a lecture. Its best moments come when the hosts admit that a standard cloud pattern is only easy until the production requirements arrive.
Made for: Mobycast is for software engineers, cloud architects, DevOps practitioners, and technical founders working with AWS and distributed systems. It especially suits listeners who want implementation details, trade-offs, and mental models rather than surface-level service summaries.
What sets it apart: The show pairs practical AWS walkthroughs with deeper distributed-systems reasoning, connecting managed-service patterns to older architectural ideas and real implementation traps. Its conversational question-and-answer style makes complex infrastructure feel inspectable without sanding off the hard parts.
A Podcast About Cloud Native Software Development, AWS, and Distributed Systems
Security broccoli comes with NAT lookup tables, and you may actually ask for seconds.
Press play if you want
Short reviews from the PlayNext desk, based on the episodes we processed.
A practical AWS architecture lesson where image downloads reveal the choices behind scale.
CloudFront becomes the hinge in Mobycast's look at a massively scalable image-hosting service. With uploads already handled, downloads bring a different architectural pressure.
A practical AWS image-hosting episode focused on why the upload path matters as much as S3 and CloudFront.
Image hosting becomes a deceptively simple AWS design problem: S3 for uploads, CloudFront for delivery, and a chain of small choices that decide whether the system holds up.
A grounded look at Leviathan's pre-cloud design and the ideas that still echo beneath DynamoDB.
The final installment turns to Leviathan, the original internet-scaled database from Chris Hickman's pre-dot-bomb company, and keeps its focus close to the architecture.
A grounded, patient look at DynamoDB's partition-level resilience, with Viathan still waiting nearby.
MobyCast's fourth installment moves DynamoDB from product-level shorthand down to the storage-node mechanics: partition-level clustering, a leader with two secondaries, heartbeats for failure detection, and Paxos for choosing a replacement.
Java conversations with architectural memory.
Cloud stories with operational weight.
Microsoft cloud security, practically explained.
Microsoft cloud security, translated into practice.
John Christensen and Chris Hickman finish their two-part Mobycast series on AWS S3 and CloudFront, moving from upload mechanics to download delivery and edge caching.
John Christensen and Chris Hickman open MobiCast with a two-part series on massively scalable image hosting using S3 and CloudFront.
Leviathan pushed the partition map into thick clients so reads could go straight to the right base server.
DynamoDB’s request router consults partition metadata, then forwards each request to the correct storage node.
DynamoDB’s partition key and sort key determine both data distribution and query speed.
Chris Hickman explains how NAT lets private machines reach the internet without exposing their addresses.
John and Chris discuss identity and access management for AWS ECS with Fargate, covering roles and best practices.
This episode introduces a series on AWS Fargate, discussing its benefits and trade-offs for running containers without managing servers.
This episode explains virtual machines and containers, comparing hypervisor implementations in cloud-native development.
John and Chris revisit their earlier discussion on virtual machines vs containers, correcting past mistakes and diving deeper into the comparison.