Closed. This question is opinion-based. It is not currently accepting answers.
Want to improve this question? Update the question so it can be answered with facts and citations by editing this post.
Closed 1 year ago.
Improve this question
Our data pipeline setup currently makes use of Docker containers running on EC2 instances and Apache Airflow as the orchestration tool. We're trying to move to MWAA for the advantages associated with a managed service(scalability, maintenance, scheduler monitoring etc...). My question is what's the best way to integrate Docker with MWAA? Currently we run our services/app via the DockerOperator on Airflow. Would we need to get Docker running on the MWAA environment? Any suggestions would be appreciated.
I would suggest to go ahead with ECS or ECS Fargate and run your Docker containers there.
MWAA supports open-source integrations with Fargate and other services, so can be easily integrated with Airflow to orchestrate
Related
Closed. This question needs to be more focused. It is not currently accepting answers.
Want to improve this question? Update the question so it focuses on one problem only by editing this post.
Closed 7 months ago.
Improve this question
I am designing a solution For real estate application with microservices (Spring Boot) on AWS. For orchestration I want to use ECS with Fargate but I am not sure of it? I am confused which will be best choice EC2 or Fargate. and also more cost efficient. please help.
If you're interested in using ECS in that case Fargate launch type would be better than EC2 because it doesn't require much maintenance and it is cost efficient.
For more information on ECS Fargate check this out https://aws.amazon.com/blogs/compute/building-deploying-and-operating-containerized-applications-with-aws-fargate/
Closed. This question is opinion-based. It is not currently accepting answers.
Want to improve this question? Update the question so it can be answered with facts and citations by editing this post.
Closed 8 months ago.
This post was edited and submitted for review 8 months ago and failed to reopen the post:
Original close reason(s) were not resolved
Improve this question
I've some questions related to AWS Cloud practitioner, DevOps, and machine learning as I'm preparing for the AWS certifications.
What is the MOST cost-effective solution to implement an image batch processing cluster on AWS? Configuring the cluster software from a generic EC2 Linux image takes 30 minutes. The application cannot run in Docker containers and must run on Amazon EC2. The batch job stores checkpoint data on an NFS and can tolerate interruptions.
Should I Use GlusterFS on EC2 instances for checkpoint data? To run the batch job, configure EC2 instances manually. When the job completes, shut down the instances manually?
And also if there is a resource with solutions to the problems related to AWS Cloud practitioner, DevOps, and machine learning would be great.
Closed. This question is opinion-based. It is not currently accepting answers.
Want to improve this question? Update the question so it can be answered with facts and citations by editing this post.
Closed 1 year ago.
Improve this question
We are trying to create a infrastructure template that can be reused for Fargate deployments.Which tool would better fit this usecase, TerraForm or cloudformation?
In my opinionated experience:
Terraform would give you a better language (HCL) and tooling (tf backends, workspaces, terragrunt, ...), and also work in other clouds and services if you need to deploy outside fargate.
CloudFormation would give you closer integration with AWS resources and services, as it is the foundation for a wide range of products. However, composing and deploying from YAML may get complex as the system grows, leading to other tools and workarounds.
You can get the "best of both" using the Terraform Cloudformation Module and so defining resources in CloudFormation but through the Terraform Tooling. Check the gitops-blueprints repo for a reference implementation.
Closed. This question is opinion-based. It is not currently accepting answers.
Want to improve this question? Update the question so it can be answered with facts and citations by editing this post.
Closed 1 year ago.
Improve this question
I need to setup Hashicorp vault on aws with cross region setup in HA. And I gotta do it with complete automation , what would be the best IAC tool , Cloudformation- I found very less documentation on this for vault setup. Or Terraform - ?
has some one achieved it by complete automation method on aws. ?
Terraform is not a Configuration Management tool, It's an IaC tool. You can use Terraform to create underline infrastructure for your Vault setup and it should not use to provision applications in the infrastructure. Of course, you can install applications in your EC2s using exec remote provisioner, but you should use provisioners as the last resort.
So I think of using Terraform for creating the infrastructure of the Vault setup. But you need to use some other tools like Ansible or Puppet to provision software in your infrastructure. Using IaC tools for configuration management will create major technical confusion in the long run.
Closed. This question is opinion-based. It is not currently accepting answers.
Want to improve this question? Update the question so it can be answered with facts and citations by editing this post.
Closed 7 months ago.
Improve this question
We are building up product which will have set of microservices communicating with each other.
We are using EKS as kubernetes service from aws. For Service mesh, There are two ways to go for it.
Install Istio over EKS cluster.
Use AWS App Mesh service with EKS.
The more I read about it, There are its own advantages of using one over other but not sure which one to go with.
Istio is more matured whereas App Mesh is able to integrate with different AWS services like XRay, Cloudwatch etc. easily.
Any inputs over this is appreciated.
This may lead to highly opinionated answers and usually this doesn't fly well on Stackoverflow. As you pointed out there isn't a specific answer other than "it depends". Your own background, organization goals, requirements etc. count more than the point opinion of individuals for this.