AWS Cloud Containers Conference - Deep Dive on Amazon ECS Capacity Providers

[@SengottuvelK]

CP2 should have 2 tasks in the slide at 14:13 It is mentioned as 3 task instead

[@cloudgreenhorn452]

Great walkthrough of the new Cluster Capacity Provider features and its strategies.

[@pradeep_kumar_ind]

This is an excellent video, because the information contained in this video is not available anywhere else. I was struggling to understand the relationship between EC2 autoscaling and Application service scaling and this video has cleared all my doubts. I do not know how I would have scaled my application without Capacity Providers. Thanks.

[@vamshikrishnatrigulla]

What is that api you have written at 27:40 ? how did the load balancer give out the task details ?

[@adamkeller1332]

Hey Vamshi! The api code is here: github.com/adamjkeller/ecsdemo-capacityproviders/blob/master/fargate/print_task_cap_providers.py What you can see in the code is I am calling the task metadata endpoint to get the information about the task running in ECS. For more information, check out the official documentation: docs.aws.amazon.com/AmazonECS/latest/developerguide/task-metadata-endpoint.html

[@ivandres73]

Great video, I learned all this with docs in 2 days. Wish I'd have found this video last week

[@andreyk6085]

Thanks for the video! What’s going to happen if there are no available spot instances of requested type in my AZ? Is my cluster going to ignore CP strategy for fargate/fargate-spot and run my tasks using on-demand only?

[@pradeep_kumar_ind]

I have one query. When I mention the capacity provider on the service update page, what about the placement strategy. Would configuration of the placement strategy has any effect on the way the capacity provider manages the tasks?

[@cloudarchitect4424]

One of the most excellent videos. Very clear and nice explanation. Thanks 👍🏻

[@ld4879]

Great video but why is this not documented anywhere else? I've spent hours using trial and error approach to make my ecs cluster scale up and down automatically :D

[@mmigliari]

The base and weight concepts are VERY confusing for a service that is supposed to make orquestration of containers EASY.

[@awssupport]

Sorry for any frustration, Maximo. I've shared your feedback internally with our AWS Events team for review. ✅🧐 ^MO

[@nexus888]

28:16 - it isn't possible to read the text on the screen. Apart from this - great presentation! :)

[@sanketshah5535]

I have question related auto scaling group and its assignment to capacity Provider. I can see cluster was created without any ec2 instances. And once we created service with ec2 type task with capacity provider it started ec2 instances. Now question is how ec2 instance profile role was attached to ec2 instance. Was that done earlier before this tutorial and what was the AMI used for that ASG, did we use ECS supported AMI? If possible can you please share that configuration of ASG (AMI used, Role how did you attach and etc..) which was used while creating Capacity Provider? Thanks in advance.

[@78nyx]

Tnx! Very clear and straight to the point o/

[@yasinlachini1791]

good knowledge