HN Theater @HNTheaterMonth

Amazon describes its internal service architecture here:

https://www.youtube.com/watch?v=dxk8b9rSKOo

Services use an internal tool called Apollo to describe the service, deploy it to EC2 VMs, and auto-scale. Apollo inspired AWS CodeDeploy and AWS AutoScaling.

Services are reachable via load balancer sets similar to AWS ELB/ALB/NLB.

You reach a service by the LB-set's DNS name.

If you don't want to use AWS or AWS-specific deployment tools, you could use Puppet Pipelines to do the same on Google Cloud or Azure. Puppet Pipelines was built by people who previously built some of those internal Amazon tools and offers similar functionality but cross-cloud.

And if you want even fewer moving parts, just go PaaS or serverless.

⬐ auslander

Never heard of Apollo, had to dig :) In short, it is a predecessor of AWS CodeDeploy.

www.allthingsdistributed.com/2014/11/apollo-amazon-deployment-engine.html

Off topic: does google.com run on GCP?

I ask because since 2010, Amazon.com has been running on AWS.

http://www.youtube.com/watch?v=dxk8b9rSKOo&t=7m32s

(Around 7m30s into video)

⬐ packetslave

Cloud runs on the exact same underlying compute, storage, and network infrastructure as search, gmail, and other Google services. The big google services are not built on top of cloud themselves (e.g. not inside GCE VMs) partially because they existed before cloud did. Some cloud services are also public versions of existing internal Google services (cloud spanner, cloud bigtable, etc)

Source: work at google, but on search, not cloud.

⬐ alberth

Are there plans for Google Services (search/gmail/etc) to move GCP anytime soon?

I'm hesitate to use GCP, given that technically Google itself isn't using GCP (totally understanding that a lift and shift of this size takes times, but it would speak volumes if you can tout that Google itself is on GCP)

Geez, oh yes. They built their platform and decided to start reselling it. That's why Netflix, Pintrest, Airbnb, Expedia, Foursquare, NASA, the MLB all use AWS.

They converted 100% in 2011 according to Jon Jenkins https://www.youtube.com/watch?v=dxk8b9rSKOo#t=449

Every developer does have the ability to deploy to production and roughly 0.001% of deployments cause an outage, but this is a company where the mean time between deployments on an average weekday is 11.6 seconds. Source: http://www.youtube.com/watch?v=dxk8b9rSKOo

Yes, it runs on EC2.

Proof: http://youtu.be/dxk8b9rSKOo?t=8m13s

Amazon runs all of the Amazon.com web servers on EC2 hosts (and has since 2010: https://www.youtube.com/watch?feature=player_detailpage&...). They've just made sure that they have enough hosts in each of the other AZs to withstand an outage.

I think you might be looking for this: http://www.youtube.com/watch?v=dxk8b9rSKOo

per the first comment: http://www.youtube.com/watch?v=dxk8b9rSKOo :)

⬐ jonjenk

Here are two other great talks by one of my friends at Amazon.

http://www.youtube.com/watch?v=UL2WDcNu_3A

http://www.youtube.com/watch?v=coNDCIMH8bk

The talk on YouTube: http://www.youtube.com/watch?v=dxk8b9rSKOo

⬐ Loic

I heavily recommend to watch the talk, 15 minutes packed with insights for the sysops. Thank you for posting the link!