OpsWorks load-based instance vs autoscale group?

Question

Does anyone know what the difference is between load-based auto-scaling and the presence of explicit auto-scaling groups in OpsWorks?

this: http://docs.aws.amazon.com/opsworks/latest/userguide/workinginstances-autoscaling-loadbased.html

vs https://aws.amazon.com/blogs/devops/auto-scaling-aws-opsworks-instances/

With load-based instances, how to add it to the target group?

Do you have multiple OpsWorks autoscale groups at the same level?

It seems to me that we are going with ALB to direct our traffic, which cannot act as an independent layer in Opsworks.

So, I will need to send requests to 1 autoscale group for one type of request, and the rest to another different scaling group.

I'm just not sure which load-based instances are perplexing without providing the default number of machines to start with.

Which one should be used to route ALB between two groups?

Answer 1

Let me share the answer for you.

Does anyone know what the difference between load-based auto-loading Scaling vs with explicit auto-scaling groups in OpsWorks is?

Automatic load-based scaling:

Amazon Opsworks Service provides you this feature, load-based auto-scaling, where you can add instances to your level in the stack and set auto-scaling policies for your configuration. Load-based scaling scales up or down instances based on the load that you configured to handle. You must set the threshold using the options and define scaling policies.

Explicit auto-scaling groups on OpsWorks:

Amazon Opsworks Service lets you add existing instances to your layer on the stack. This means that you can configure the autoscale trigger configuration and set the scale and zoom out of events based on the load. Then create an Autoscaling group and run instances in it. You can then go to Opsworks and add these existing instances to your layer on the stack. Therefore, when the load increases or decreases more or less than the set threshold, the Autoscaling group processes the scaling.

With load-based instances, how to add one to the target group?

Once you have ready-to-download instances that you started directly from load-based automatic scaling in Opsworks or explicitly using the Auto Scaling groups on OpsWork, you can go to the application download manager in the EC2 console and configure using the necessary configurations and then register the load-based instances you just created with ALB in the TAB Goal Registry .

Do you have multiple OpsWorks autoscale groups at the same level?

Yes , you can have multiple auto-scaling groups at the same level as OpsWorks.

Which one should be used to route ALB between two groups?

You can use any of the groups.

so that you can send requests to 1 auto-scaling group for one type of request, and the rest to another auto-scaling group.

Refer to autosave once.

I'm just not sure which load based instances

Load-based instances are instances that are configured using a load-based scaling configuration. You must set the threshold, configuration, and events to determine when to scale and scale. EX: Suppose if you initially have 5 instances, and if you want your application to run even on your load increase, to minimize downtime, you will configure the autoscale configuration so that if the processor uses average instances increased by more than 70% 2 more copies. You can zoom in and out on many other factors.

Hope this helps :)

Answer 2

OpsWorks is a configuration management tool that Chef uses to customize your infrastructure. OpsWorks takes a different approach when it comes to scaling than the auto-scaling group.

Unlike the auto-scaling group, you have these instances predefined on your OpsWorks stack (level), and they start when a certain indicator (threshold) is triggered ( CloudWatch : CPU, memory, loading ... etc.).

OpsWorks will not create (create) any new instances, but will only be able to run the instances that you previously created and install them as instances based on the load. It is also only available for OpsWorks and cannot be used for any other services outside of OpsWorks.

Auto-scaling AWS EC2 can actually launch a very large number of instances (instances that do not need to be created in advance) to the AWS environment, and the same as scaling based on OpsWorks load can be caused by CloudWatch signals (CPU, memory, loading .. . etc.).

Autoscaling by default is not available on OpsWorks, and there is no way to build an autoscaling group associated with your OpsWorks stack, but it is possible with a little work. Read about it here .

Answer 3

I like the answers that come to this question, but I think they both missed something.

I will try to be very concise:

The main use case for automatic scaling groups is to use them with load balancers. In this case, you should still use the additional configuration to register your load-balanced instances in the OpsWorks stack. Check it out. Basically you will need Lambda and SNS for shutdown events and a user data script when new instances are created through the Auto Scaling Group.

In scenarios when you:

1. Use OpsWorks as a Configuration Manager;

2. And AWS Elastic Load Balancing is not required (maybe you use some of them, for example, back office workloads that can work independently or use their own load balancing, say, through Nginx);

... you can simply use load-based OpsWorks instances and set up scaling in one place.