Add reason label to total_reload metric

[LorcanMcVeigh]

Proposed changes

Add a label reason to the prometheus metric total_reloads which specifies if the reload was caused by a endpoint update or another reason

Checklist

Before creating a PR, run through this checklist and mark each as complete.

• I have read the CONTRIBUTING doc
• I have added tests that prove my fix is effective or that my feature works
• I have checked that all unit tests pass after adding my changes
• I have updated necessary documentation
• I have rebased my branch onto master
• I will ensure my PR is targeting the master branch and pulling from my branch from my own fork

[pleshakov]

@LorcanMcVeigh
Please see my comments and suggestions

Additionally, after you start the IC, until it gets reloaded, you will not get metrics for one or both labels. For "other" label, as the first reload happens quickly after the start, it is hard to notice its absence. However, for the "endpoints" label, you can see its absence. For example, here the IC was reloaded but no endpoints update happened yet:

# HELP nginx_ingress_controller_nginx_reloads_total Number of successful NGINX reloads
# TYPE nginx_ingress_controller_nginx_reloads_total counter
nginx_ingress_controller_nginx_reloads_total{class="nginx",reason="other"} 2

Note that the metric nginx_ingress_controller_nginx_reloads_total{class="nginx",reason="endpoint"} is not present here at all.

The solution is to initialize those labels before the IC starts. Could you add that?

[LorcanMcVeigh]

@LorcanMcVeigh
Please see my comments and suggestions

Additionally, after you start the IC, until it gets reloaded, you will not get metrics for one or both labels. For "other" label, as the first reload happens quickly after the start, it is hard to notice its absence. However, for the "endpoints" label, you can see its absence. For example, here the IC was reloaded but no endpoints update happened yet:

# HELP nginx_ingress_controller_nginx_reloads_total Number of successful NGINX reloads
# TYPE nginx_ingress_controller_nginx_reloads_total counter
nginx_ingress_controller_nginx_reloads_total{class="nginx",reason="other"} 2

Note that the metric nginx_ingress_controller_nginx_reloads_total{class="nginx",reason="endpoint"} is not present here at all.

The solution is to initialize those labels before the IC starts. Could you add that?

could you possibly tell me how you achieved this? I haven't been able to replicate this

[pleshakov]

@LorcanMcVeigh

could you possibly tell me how you achieved this? I haven't been able to replicate this

yep. So just deploy an Ingress Controller. Without deploying any backends and Ingress/VS resources.

[pleshakov]

👍