v1.4.1

Fix issue where sombra instances using HTTP behind the load balancer did not have the correct ALB target group type set

v1.4.0

Allow customizing load balancer health check protocol

v1.3.3

Fixes a typo in our main.tf file that would result in the module not being able to pick up some OAUTH config.

v1.3.2

Adds flexibility for oauth2 configuration from separate URLs

v1.3.1

Add resource label inputs for ALB based scaling.

v1.3.0

Add support for AWS Application Auto-scaling. This is intended to replace the faulty v1.2.0 release, which did not actually have the changes we wanted.

Support idle_timeout on Load Balancers

Allows changing the idle_timeout for the load balancer for the sombra service.

v1.1.0

Added TLS v1.2 as the minimum accepted TLS version for improved security

v1.0.0

Up until now, the only available control on this module for cpu/memory were the cpu and memory variables. These variables applied to the ECS service, which made sense as the previous assumption was that the sombra container would always have fixed CPU and memory (512 vCPU, and 1Gb memory).

This PR adds the ability to customize the container-level CPU and memory, and also increases the default RAM usage to 2Gb to coincide with recent changes to our node runtime environment plugins that raised our memory consumption.

This can be considered a breaking change, as it is possible (even likely) that clients using this module have their total service cpu/mem limits set fairly close to how much actual cpu/mem they use, and the raise in default RAM usage on the sombra container may push them over their memory limit they had previously set.

To upgrade: Just run an normal apply on the new version, ensuring that your memory variable value is greater than the sum of all container memory values you add to the Task plus the memory of the sombra container.

Allow Backend-Lambda runners to access Sombra instances

Merge pull request #40 from transcend-io/dipack/update-lambda-ip-redux

Add fallback Lambda VPC Elastic IP.