NetApp Harvest configuration

[chaithi88]

Hi,
I am working on setting up Netapp harvest using Prometheus and I am running into an issue where I am not able to see the NetApp filer included in the pollers is not discovered on the Prometheus targets. Can you please help review config and let me know where/what the issue is? Appreciate your help. Thanks

root@ntap-harvest-poc: ~# cat /opt/harvest/harvest.yml
version: 1

logging:
level: debug
file: /opt/harvest/harvest.log
Exporters:
prometheus:
exporter: Prometheus
local_http_addr: 0.0.0.0
port: 12990

Defaults:
collectors:
- Zapi
- ZapiPerf
- Rest
- RestPerf
- Ems
use_insecure_tls: true

Pollers:
unix:
datacenter: local
addr: localhost
collectors:
- Unix
exporters:
- prometheus

cluster-01:
datacenter: IAD1
addr: 10.x.x.x
auth_style: basic_auth
username: admin
password: xxxxxxx
use_insecure_tls: true
exporters:
- prometheus
schedule:
1: 1m # example to run every minute

[cgrinds]

Not sure how you installed Harvest, but can you share your log file?
It should tell us what’s going on.
https://netapp.github.io/harvest/nightly/help/log-collection/

We have a bin/harvest doctor command that is handy to validate your harvest.yml file. It will point out common errors.

If I run “harvest doctor” with your config file it points out a problem with your Prometheus exporters, which I’ve pasted here:

Error: Multiple pollers export to the same PromPort
  Each poller should export to a unique Prometheus exporter or use PortRange. Change the following pollers to use unique exporters:
  pollers [unix, cluster-01] export to the same static PrometheusExporter: [prometheus]

When using a Prometheus exporter, each poller needs its own unique port. In your case, you are attempting to make two different pollers (unix and cluster-01) use the same port. That won’t work, and the log files should say that.

You can create another Prometheus exporter or comment out the unix poller. Below, I've pasted a new harvest.yml with an additional Prometheus exporter that should work for you. If you decide to use the attached harvest.yml, you’ll need to update your Prometheus scrape target to include the unix poller too.

Exporters:
    prometheus1:
        exporter: Prometheus
        local_http_addr: 0.0.0.0
        port: 12990
    prometheus2:
        exporter: Prometheus
        local_http_addr: 0.0.0.0
        port: 12991   
Defaults:
    collectors:
        - Zapi
        - ZapiPerf
        - Rest
        - RestPerf
        - Ems
    use_insecure_tls: true
Pollers:
    unix:
        datacenter: local
        collectors:
            - Unix
        exporters:
            - prometheus2
    cluster-01:
        datacenter: IAD1
        addr: -REDACTED-
        auth_style: basic_auth
        username: -REDACTED-
        password: -REDACTED-
        use_insecure_tls: true
        exporters:
            - prometheus1

[cgrinds]

Closing. @chaithi88 confirmed they're all set in email.

Summary of issues:

• static_configs target port mismatch between harvest.yml and prometheus.yml
• missing permissions for Harvest user
• bin/harvest doctor pointed out a problem with harvest.yml - the issue was multiple pollers were exporting to the same PromPort