coldbrewcloud / coldbrew-cli Goto Github PK

Mostly output text colors

Is there any way to create a task definition that contains multiple containers using the cli?

I currently use the ecs-cli to bring up a service using a docker-compose file but then have a big build script that has to create all the dependencies such as target groups etc. Using this cli would make things so much easier however we do have to run multiple containers within a single service.

When cluster name is not provided:

1. If application configuration is found (based on global flags), use the cluster name from the configuration file
2. otherwise, ask user to enter cluster name

Test Configuration (INPUT of each test units)

• test app directory
• test cluster name
• test app configuration file

Test Procedure:

• Create cluster
• Verify all expected resources are properly created and configured
• Create app
• Verify all expected resources are properly created and configured
• Delete app
• Verify all expected resources are properly deleted and updated
• Delete cluster
• Verify all expected resources are properly deleted

hello! looking forward to Jay's blogpost. i tried deploying https://github.com/cefjoeii/mern-crud with coldbrew but i think the mongodb config led me to be unable to serve the page (i just get 503'ed - http://tutorial-nodejs-elb-1262210809.us-east-1.elb.amazonaws.com/). Its hard to tell (if its the mdb config's fault) without being able to get logs out of coldbrew to see if my app is crashing somewhere. does it make sense to have a coldbrew logs command to give more color to the coldbrew status command?

Adding --app-name and --cluster-name flags to status and delete command so they can run without the configuration file (e.g. coldbrew.conf).

Error: Failed to create ELB Load Balancer [xxx-test-elb]: InvalidConfigurationRequest: A load balancer cannot be attached to multiple subnets in the same Availability Zone.

Please, let me know any who have same situation.
Thanks in advance.

For the service-level changes (desired tasks), it doesn't have to create a new ECS task definition revision.

Update build script and/or CI config for versioned binary packages

How would I pass environment variables to my node application?

I see how you can pass key value pairs into the config:
https://github.com/coldbrewcloud/coldbrew-cli/wiki/Configuration-File#env

But this means you are committing sensitive data to the root of your repo. Is there a way to pull from a .env file outside of the root?

I have a question I just tested with a meteor application deployment and everything to walk properly it's a great job, but what should set up on ELB (ALB) sticky session for the websocket?

If yes can be put an option to add it directly in the configuration?

This is a great tool, just to let you know that ECS agent version has changed to version 1.13.1, and version 1.4.1 is not present in releases

Adding this argument doesn't seem to have any effect (using the API as in https://github.com/coldbrewcloud/coldbrew-cli/wiki/CLI-Command:-cluster-create)

command finishes if i delete cluster from console.

bash-3.2$ coldbrew cluster-delete demo-platespace
Determining AWS resources that need to be deleted...
  ECS Cluster: coldbrew-demo-platespace
  IAM Role for ECS Services: coldbrew-demo-platespace-ecs-service-role
  EC2 Launch Configuration for ECS Container Instances: coldbrew-demo-platespace-lc
  EC2 Auto Scaling Group for ECS Container Instances: coldbrew-demo-platespace-asg
  IAM Instance Profile for ECS Container Instances: coldbrew-demo-platespace-instance-profile
  EC2 Security Group for ECS Container Instances: coldbrew-demo-platespace-instance-sg

> Do you want to delete these resources? [y/N]: y

[*] Terminating instances in EC2 Auto Scaling Group [coldbrew-demo-platespace-asg]... (this may take long)
[-] Deleting EC2 Auto Scaling Group [coldbrew-demo-platespace-asg]... (this may take long)
[-] Deleting EC2 Launch Configuration [coldbrew-demo-platespace-lc]...
[-] Deleting IAM Instance Profile [coldbrew-demo-platespace-instance-profile]...
[-] Deleting EC2 Security Group [coldbrew-demo-platespace-instance-sg]...
[-] Deleting ECS Cluster [coldbrew-demo-platespace]...

Error: Failed to delete ECS Cluster [coldbrew-demo-platespace-ecs-service-role]: ClusterContainsServicesException: The Cluster cannot be deleted while Services are active.
	status code: 400, request id: 87dbb57b-7ba6-11e7-8f63-4549790bf4f0

--aws-secrte-key
must be
--aws-secret-key

On this wiki page

Every command I try gives me this error:

panic: interface conversion: io.Writer is *colorable.Writer, not *os.File

Running on windows 10.

delete and cluster-delete commands can ask user if they want to untagged (possibly manually created) resources.

It would be helpful to give some explanation of the added value of coldbrew over native ECS.

Challenge: when scaling in, instances with none or less number of tasks running should be prioritized (termination policy).

And it causes the instances fail to connect to ECS cluster (because they don't have access).

Guessing this is some delays for the newly created Instance Profile become fully available, but, not sure. Need to investigate.

I guess it's just not printing out its progress output, but, it's still bad experience. Need to fix.

I tried going through the example Go tutorial:
https://github.com/coldbrewcloud/tutorial-echo

I installed via go get and go install, so using the master branch on the CLI.

$ coldbrew status
Application
  Name: echo
  Cluster: tutorial
AWS
  Region: us-west-2
  VPC: vpc-69db9410
  Subnets: subnet-0b0ed840 subnet-05df687c subnet-87e473dd
ECS
  ECS Cluster: coldbrew-tutorial
  ECS Service: echo
  ECS Task Definition: echo:1
  Tasks (current/desired/pending): 0/2/0
Container Definition
  Name: echo
  Image: 272340174925.dkr.ecr.us-west-2.amazonaws.com/coldbrew/echo:latest
  CPU: 0.10
  Memory: 100m
  Port Mapping (protocol:container:host): tcp:8888:0
Load Balancer
  Container Port: 8888
  ELB Target Group: echo-elb-tg
  ELB Load Balancer: echo-elb
    Scheme: internet-facing
    State: active 
    Endpoint: http://echo-elb-216009805.us-west-2.elb.amazonaws.com:80

The containers are never launched on the cluster. In the ECS control panel it gives a warning that the ecs agent is out of date (1.14.4 vs 1.17.3), and links to https://docs.aws.amazon.com/AmazonECS/latest/developerguide/ecs-agent-update.html

Option 1. configure tasks to send output logs to CloudWatch Logs, then CLI retrieves logs from there.

http://docs.aws.amazon.com/AmazonECS/latest/developerguide/private-auth.html

A utility function to create a new EC2 Key Pair.

And/or multiple load balancer

Any plan to support other platforms in the future? DigitalOcean?

Scaling In

E.g. need to remove container instance C1 which is running tasks for S1 and S2.

• Deregister C1 from ECS cluster (use --force option to keep tasks running)
• Tasks for S1 and S2 are still running in C1, but, ECS cluster will try to start tasks for S1 and S2 in other container instances.
• Once services S1 and S2 become stable status (desired tasks == running tasks), stop all tasks in C1.
• Detach C1 from auto scaling group (use --should-decrement-desired-capacity)
• Terminate C1.
  For multiple container instances, simply repeat these steps for each instances.

Scaling Out

E.g. need to add N container instances.

• Increment desired capacity of auto scaling group by N.

If I am an IAM user without full root privs, what IAM user access do I need to create all parts of a coldbrew created ecs cluster

correct "Balanacer"

Correct spelling is "balancer"

https://github.com/coldbrewcloud/coldbrew-cli/blob/master/commands/deploy/aws_elb.go