AWS re:Invent 2015 | (CMP406) Amazon ECS at Coursera
2 796
54.8
Amazon Web Services776 тыс
Следующее
12.10.15 – 21 99856:26Популярные
Опубликовано 12 октября 2015, 20:18
Coursera has helped millions of students learn computer science through MOOCs ranging from Introduction to Python, to state-of-the-art Functional-Reactive Programming in Scala. Our interactive educational experience relies upon an automated grading platform for programming assignments. But, because anyone can sign up for a course on Coursera for free, our systems must defend against arbitrary code execution.
Come learn how Coursera uses AWS services such as Amazon EC2 Container Service (ECS), and Amazon Virtual Private Cloud (VPC) to power a defense-in-depth strategy to secure our infrastructure against bad actors. We have modified the Amazon ECS Agent to support security layers including kernel privilege de-escalation, and enabling mandatory access control systems. Additionally, we post-process uploaded grading container images to defang binaries.
At the core of automated grading is a general-purpose near-line & batch scheduling and execution microservice built on top of the Amazon ECS APIs. We use this flexible system to power a variety of internal services across the company including data exports for instructors, course announcement emails, data reconciliation jobs, and more.
In this session, we detail aspects of our success from implementing Docker and Amazon ECS in production, providing ideas for your own scheduling, execution and hardening requirements.
Come learn how Coursera uses AWS services such as Amazon EC2 Container Service (ECS), and Amazon Virtual Private Cloud (VPC) to power a defense-in-depth strategy to secure our infrastructure against bad actors. We have modified the Amazon ECS Agent to support security layers including kernel privilege de-escalation, and enabling mandatory access control systems. Additionally, we post-process uploaded grading container images to defang binaries.
At the core of automated grading is a general-purpose near-line & batch scheduling and execution microservice built on top of the Amazon ECS APIs. We use this flexible system to power a variety of internal services across the company including data exports for instructors, course announcement emails, data reconciliation jobs, and more.
In this session, we detail aspects of our success from implementing Docker and Amazon ECS in production, providing ideas for your own scheduling, execution and hardening requirements.
Свежие видео
3 дня – 3 3431:18
4 дня – 1 6669:21
5 дней – 5730:58
9 дней – 5 0041:45
12 дней – 688 8256:21Случайные видео
183 дня – 3 285 74211:07
190 дней – 35 5280:18
320 дней – 4 4820:59
10.05.21 – 18 7553:45
08.02.21 – 2 9730:39
22.03.16 – 4 7093:20
2 дня – 2 1921:02
3 дня – 1 2260:50
4 дня – 7 5353:24
4 дня – 988 72717:54
5 дней – 23113:04
6 дней – 21 22841:38
