You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
# Gather the metrics from the prometheus deployed by Rancher
11
-
For the control plane metrics, the services are not created by default. To get them in the Prometheus server you have to create new services and seviceMonitors. To gather that metrics with the Sysdig agent, you have to create the rules to filter them, and then federate the metrics with the agent itself.
12
-
13
-
You can either follow the steps given below or download the script and execute:
14
-
```sh
15
-
sh installation.sh
16
-
```
17
-
And then apply the configuration changes for the `sysdig-agent`.
6
+
For the control plane metrics, the services are not created by default. To get them in the Prometheus server you have to create new services and seviceMonitors.
In order to collect the metrics you have to enable `promscrape_fastproto` to do. Therefore, ensure that your `dragent.yaml` includes the following values:
30
-
```yaml
31
-
metrics_excess_log: true
32
-
k8s_cluster_name: yourClusterName
33
-
10s_flush_enable: true
34
-
app_checks_enabled: false
35
-
use_promscrape: true
36
-
new_k8s: true
37
-
promscrape_fastproto: true
38
-
prometheus:
39
-
enabled: true
40
-
prom_service_discovery: true
41
-
log_errors: true
42
-
max_metrics: 200000
43
-
max_metrics_per_process: 200000
44
-
max_tags_per_metric: 100
45
-
ingest_raw: true
46
-
ingest_calculated: false
47
-
```
48
-
49
-
See the example `sysdig-agent.yaml` file given below:
50
-
51
-
With the promscrape v2 enabled you can scrape the etcd. To do so, ensure that your `sysdig-agent.yaml` includes the prometheus job as given below:
0 commit comments