Try OpenEdge Now
skip to main content
Administration Guide
Load balancing options for PAS for OpenEdge : Amazon load balancing : Configuring Elastic Load Balancing and Auto Scaling
 

Configuring Elastic Load Balancing and Auto Scaling

This topic is an overview of configuring Amazon Elastic Load Balancing and Auto Scaling as it applies to PAS for OpenEdge. For more information about the Amazon products, see https://aws.amazon.com/documentation/elastic-load-balancing/ and https://aws.amazon.com/documentation/autoscaling/.
The following table lists the relevant terms applied to the components of these products.
Table 7. Elastic Load Balancing and Auto Scaling components
Component
Description
Virtual Private Cloud (VPC)
A private network of your application machines, which is your internal, "safe" network away from the internet. Access can be limited to specific incoming IP addresses or ranges.
Amazon Machine Image (AMI)
A copy of a virtual machine that, when launched, will automatically start the processes that support the load balanced application.
Elastic Load Balancer (ELB)
An internet facing load balancer that redirects work to the instances running in a VPC. Monitors instance health and redirects using a proprietary cookie (AWS_ELB).
Launch Configuration
A definition of the configuration of the AMI you want to launch automatically when scaling up.
Auto Scaling Groups
A definition of a group of AMI's that specifies their ELB and the scaling policy (how and when scaling up or down occurs).
The basic steps for creating an Elastic Load Balancing and Auto Scaling configuration are:
1. Create a VPC (or use the default) which includes:
*defining a network mask range for your private network
*defining a security group that limits access to the private network
2. Create the AMIs.
Each AMI must have a PAS for OpenEdge instance installed, and the instance must be configured to start when the operating system launches. Note that the instances can be configured to use the same ports since they are running on different hosts and therefore will have unique IP addresses in the VPC.
3. Create an ELB, enabling:
*an internet facing scheme
*port forwarding (for example, forward port 80 to the PAS for OpenEdge port 8810)
*sticky sessions (using the AWS_ELB cookie) or cookie timeout.
*cross-zone load balancing
*log timing
*idle timeout
*a Health Check
4. Create a Launch Configuration that specifies:
*an AMI
*the instance size (CPUs, memory, and disk space)
*the VPC
*the security group
5. Create an Auto Scaling Group that specifies:
*the ELB
*the VPC
*the maximum and minimum number of copys of the AMI that can be launched
*a Health Check configuration
*the scaling policy (what metric threshold causes scaling up or down)
*the tag to apply at instance launch
Note: The preceding steps are just an overview of the procedures required to implement Amazon load balancing. See the Amazon documentation for more complete information.