Remove VTS from the ingress controller
This commit is contained in:
Manuel de Brito Fontes
Manuel Alejandro de Brito Fontes
parent
9c6201b79a
commit
63b38e1c21
18 changed files with 1 additions and 250 deletions
|
|
@ -42,6 +42,3 @@ $ kubectl exec nginx-ingress-controller-v1ppm cat /etc/nginx/nginx.conf
|
|||
}
|
||||
....
|
||||
```
|
||||
|
||||
|
||||
|
||||
|
|
|
|||
Binary file not shown.
|
Before Width: | Height: | Size: 59 KiB |
|
|
@ -1,103 +0,0 @@
|
|||
# Custom VTS metrics with Prometheus
|
||||
|
||||
This example aims to demonstrate the deployment of an nginx ingress controller and use a ConfigMap to enable [nginx vts module](https://github.com/vozlt/nginx-module-vts
|
||||
) to export metrics in prometheus format.
|
||||
|
||||
## vts-metrics
|
||||
|
||||
Vts-metrics export NGINX metrics. To deploy all the files simply run `kubectl apply -f nginx`. A deployment and service will be
|
||||
created which already has a `prometheus.io/scrape: 'true'` annotation and if you added
|
||||
the recommended Prometheus service-endpoint scraping [configuration](https://raw.githubusercontent.com/prometheus/prometheus/master/documentation/examples/prometheus-kubernetes.yml),
|
||||
Prometheus will scrape it automatically and you start using the generated metrics right away.
|
||||
|
||||
## Custom configuration
|
||||
|
||||
```console
|
||||
apiVersion: v1
|
||||
data:
|
||||
enable-vts-status: "true"
|
||||
kind: ConfigMap
|
||||
metadata:
|
||||
name: nginx-configuration
|
||||
namespace: ingress-nginx
|
||||
labels:
|
||||
app: ingress-nginx
|
||||
```
|
||||
|
||||
```console
|
||||
$ kubectl apply -f nginx-vts-metrics-conf.yaml
|
||||
```
|
||||
|
||||
## Result
|
||||
|
||||
Check whether the ingress controller successfully generated the NGINX vts status:
|
||||
|
||||
```console
|
||||
$ kubectl exec nginx-ingress-controller-873061567-4n3k2 -n ingress-nginx cat /etc/nginx/nginx.conf|grep vhost_traffic_status_display
|
||||
vhost_traffic_status_display;
|
||||
vhost_traffic_status_display_format html;
|
||||
```
|
||||
|
||||
### NGINX vts dashboard
|
||||
|
||||
The vts dashboard provides real time metrics.
|
||||
|
||||
|
||||
|
||||
Because the vts port it's not yet exposed, you should forward the controller port to see it.
|
||||
|
||||
```console
|
||||
$ kubectl port-forward $(kubectl get pods --selector=k8s-app=nginx-ingress-controller -n ingress-nginx --output=jsonpath={.items..metadata.name}) -n ingress-nginx 18080
|
||||
```
|
||||
|
||||
Now open the url [http://localhost:18080/nginx_status](http://localhost:18080/nginx_status) in your browser.
|
||||
|
||||
### Prometheus metrics output
|
||||
|
||||
NGINX Ingress controller already has a parser to convert vts metrics to Prometheus format. It exports prometheus metrics to the address `:10254/metrics`.
|
||||
|
||||
```console
|
||||
$ kubectl exec -ti -n ingress-nginx $(kubectl get pods --selector=k8s-app=nginx-ingress-controller -n kube-system --output=jsonpath={.items..metadata.name}) curl localhost:10254/metrics
|
||||
ingress_controller_ssl_expire_time_seconds{host="foo.bar.com"} -6.21355968e+10
|
||||
# HELP ingress_controller_success Cumulative number of Ingress controller reload operations
|
||||
# TYPE ingress_controller_success counter
|
||||
ingress_controller_success{count="reloads"} 3
|
||||
# HELP nginx_bytes_total Nginx bytes count
|
||||
# TYPE nginx_bytes_total counter
|
||||
nginx_bytes_total{direction="in",ingress_class="nginx",namespace="",server_zone="*"} 3708
|
||||
nginx_bytes_total{direction="in",ingress_class="nginx",namespace="",server_zone="_"} 3708
|
||||
nginx_bytes_total{direction="out",ingress_class="nginx",namespace="",server_zone="*"} 5256
|
||||
nginx_bytes_total{direction="out",ingress_class="nginx",namespace="",server_zone="_"} 5256
|
||||
```
|
||||
|
||||
### Customize metrics
|
||||
|
||||
The default [vts vhost key](https://github.com/vozlt/nginx-module-vts#vhost_traffic_status_filter_by_set_key) is `$geoip_country_code country::*` that expose metrics grouped by server and country code. The example below show how to have metrics grouped by server and server path.
|
||||
|
||||
|
||||
|
||||
## NGINX custom configuration ( http level )
|
||||
|
||||
```
|
||||
apiVersion: v1
|
||||
kind: ConfigMap
|
||||
data:
|
||||
enable-vts-status: "true"
|
||||
vts-default-filter-key: "$server_name"
|
||||
...
|
||||
```
|
||||
|
||||
## Customize ingress
|
||||
|
||||
```
|
||||
apiVersion: extensions/v1beta1
|
||||
kind: Ingress
|
||||
metadata:
|
||||
annotations:
|
||||
nginx.ingress.kubernetes.io/vts-filter-key: $uri $server_name
|
||||
name: ingress
|
||||
```
|
||||
|
||||
## Result
|
||||
|
||||
|
||||
Binary file not shown.
|
Before Width: | Height: | Size: 969 KiB |
Binary file not shown.
|
Before Width: | Height: | Size: 451 KiB |
Binary file not shown.
|
Before Width: | Height: | Size: 244 KiB |
|
|
@ -1,9 +0,0 @@
|
|||
apiVersion: v1
|
||||
data:
|
||||
enable-vts-status: "true"
|
||||
kind: ConfigMap
|
||||
metadata:
|
||||
name: nginx-configuration
|
||||
namespace: ingress-nginx
|
||||
labels:
|
||||
app: ingress-nginx
|
||||
|
|
@ -16,7 +16,6 @@ Customization | [Custom DH parameters for perfect forward secrecy](customization
|
|||
Customization | [Custom errors](customization/custom-errors/README.md) | serve custom error pages from the default backend | Intermediate
|
||||
Customization | [Custom headers](customization/custom-headers/README.md) | set custom headers before sending traffic to backends | Advanced
|
||||
Customization | [Custom upstream check](customization/custom-upstream-check/README.md) | TODO | TODO
|
||||
Customization | [Custom VTS metrics with Prometheus](customization/custom-vts-metrics-prometheus/README.md) | TODO | TODO
|
||||
Customization | [External authentication with response header propagation](customization/external-auth-headers/README.md) | TODO | TODO
|
||||
Customization | [Sysctl tuning](customization/sysctl/README.md) | TODO | TODO
|
||||
Features | [Rewrite](rewrite/README.md) | TODO | TODO
|
||||
|
|
|
|||
|
|
@ -44,10 +44,6 @@ The following table shows a configuration option's name, type, and the default v
|
|||
|[disable-ipv6-dns](#disable-ipv6-dns)|bool|false|
|
||||
|[enable-underscores-in-headers](#enable-underscores-in-headers)|bool|false|
|
||||
|[ignore-invalid-headers](#ignore-invalid-headers)|bool|true|
|
||||
|[enable-vts-status](#enable-vts-status)|bool|false|
|
||||
|[vts-status-zone-size](#vts-status-zone-size)|string|"10m"|
|
||||
|[vts-sum-key](#vts-sum-key)|string|"*"|
|
||||
|[vts-default-filter-key](#vts-default-filter-key)|string|"$geoip_country_code country::*"|
|
||||
|[retry-non-idempotent](#retry-non-idempotent)|bool|"false"|
|
||||
|[error-log-level](#error-log-level)|string|"notice"|
|
||||
|[http2-max-field-size](#http2-max-field-size)|string|"4k"|
|
||||
|
|
@ -240,32 +236,6 @@ Enables underscores in header names. _**default:**_ is disabled
|
|||
Set if header fields with invalid names should be ignored.
|
||||
_**default:**_ is enabled
|
||||
|
||||
## enable-vts-status
|
||||
|
||||
Allows the replacement of the default status page with a third party module named [nginx-module-vts](https://github.com/vozlt/nginx-module-vts).
|
||||
_**default:**_ is disabled
|
||||
|
||||
## vts-status-zone-size
|
||||
|
||||
Vts config on http level sets parameters for a shared memory zone that will keep states for various keys. The cache is shared between all worker processes. _**default:**_ 10m
|
||||
|
||||
_References:_
|
||||
[https://github.com/vozlt/nginx-module-vts#vhost_traffic_status_zone](https://github.com/vozlt/nginx-module-vts#vhost_traffic_status_zone)
|
||||
|
||||
## vts-default-filter-key
|
||||
|
||||
Vts config on http level enables the keys by user defined variable. The key is a key string to calculate traffic. The name is a group string to calculate traffic. The key and name can contain variables such as $host, $server_name. The name's group belongs to filterZones if specified. The key's group belongs to serverZones if not specified second argument name. _**default:**_ $geoip_country_code country::*
|
||||
|
||||
_References:_
|
||||
[https://github.com/vozlt/nginx-module-vts#vhost_traffic_status_filter_by_set_key](https://github.com/vozlt/nginx-module-vts#vhost_traffic_status_filter_by_set_key)
|
||||
|
||||
## vts-sum-key
|
||||
|
||||
For metrics keyed (or when using Prometheus, labeled) by server zone, this value is used to indicate metrics for all server zones combined. _**default:**_ *
|
||||
|
||||
_References:_
|
||||
[https://github.com/vozlt/nginx-module-vts#vhost_traffic_status_display_sum_key](https://github.com/vozlt/nginx-module-vts#vhost_traffic_status_display_sum_key)
|
||||
|
||||
## retry-non-idempotent
|
||||
|
||||
Since 1.9.13 NGINX will not retry non-idempotent requests (POST, LOCK, PATCH) in case of an error in the upstream server. The previous behavior can be restored using the value "true".
|
||||
|
|
|
|||
|
|
@ -2,10 +2,3 @@
|
|||
|
||||
The [ngx_http_stub_status_module](http://nginx.org/en/docs/http/ngx_http_stub_status_module.html) module provides access to basic status information.
|
||||
This is the default module active in the url `/nginx_status` in the status port (default is 18080).
|
||||
|
||||
This controller provides an alternative to this module using the [nginx-module-vts](https://github.com/vozlt/nginx-module-vts) module.
|
||||
To use this module just set in the configuration configmap `enable-vts-status: "true"`.
|
||||
|
||||
|
||||
|
||||
To extract the information in JSON format the module provides a custom URL: `/nginx_status/format/json`
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue