-
Notifications
You must be signed in to change notification settings - Fork 96
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Make metric port configurable #101
Comments
@iffyio I was going to dig into this, but found a couple of interesting things: Within our proxy config docs, we have a We do however have an Looking at Envoy - they do have a /stats that comes off the same port as the admin interface -- are we happy doing the same thing, or is there any reason we would want metrics to run off a different port to the admin port? I am thinking that if we ever add paths to the admin port that allow mutation of the proxy, it would be more secure to have metrics be on a separate port (just digging into some security related Envoy discussion). So I think I would be pro-making it a separate port from everything else, and can be secured (or in the case of metrics, be more open) separately from anything else. If we agree on that, I'll do the bits to flesh out the model, and get that integrated 👍 |
Yeah the doc isn't correct looks more like. I think we can run metrics on the same admin port as envoy currently does and update to the docs to match that |
Okay cool - it seems like it'll be fine. We can also do a TLS cert or other auth methods down the line if we decide we need more access control. Working on this now. |
I should say - as part of this work, I'll move the hyper server into it's own "Admin" module, and come up with a strategy that allows other modules to plug in. |
Update the model and documentation for the Admin interface to match, and provide defaulting. Next will be the implementation of the `Admin` functionality, and pulling it out of `Metrics`. Work on #101
This pulls the `hyper` server out of the `Metrics` module and move it into its own `Admin` module that handles metrics, and in the future, the health/liveness endpoint as well. Closes #101
This pulls the `hyper` server out of the `Metrics` module and move it into its own `Admin` module that handles metrics, and in the future, the health/liveness endpoint as well. Closes #101
Just ran into this when trying to run a local demo - couldn't spin up two instances of quilkin locally, as both were trying to bind to 9091.
Would be handy to make this configurable in the yaml.
The text was updated successfully, but these errors were encountered: