Introduction
There are a number of application configuration settings available through replicatedctl that may need to be adjusted, in order to tune overall performance of a Terraform Enterprise instance for company needs.
As an example, the capacity_concurrency setting can be modified to adjust the maximum number of concurrent runs allowed at any given time.
In an Active/Active setup, it is possible to accomplish the above without incurring downtime due to application restart.
Recommendation
- Establish an SSH connection with one of the nodes in the Active/Active configuration.
- Run a node-drain to gracefully remove the node from service:
tfe-admin node-drain
- Stop the application on the node:
replicatedctl app stop
- Ensure the application has stopped:
watch replicatedctl app status # ensure "State" is "stopped"
- Adjust the setting of interest with
replicatedctl app set <setting> --value <desired_value>
- Restart the application:
replicatedctl app start # ensure "State" is "started"
- Ensure the application has started:
watch repliactedctl app status
- Repeat for other nodes in the Active/Active configuration.
Additional Information
- To view settings that can be modified, use replicatedctl app-config export
-
Information on the tfe-admin node-drain command: https://support.hashicorp.com/hc/en-us/articles/4406210845331-How-to-stop-Terraform-Enterprise-active-active-node