Harvest: Metric collection fails with error: context deadline exceeded (Client.Timeout)

[dusky niche]

The ONTAP cluster is running in a VM. They have similar 7 cluster where harvest is collecting metrics fine with similar configuration
The harvest is running in a POD inside Openshift Cluster.

However when we try metric collection from this cluster it shows below error.

time=2026-01-15T0754.459Z level=WARN source=poller.go:1527 msg="gather cluster info" Poller=iaas-dev-netapp remote="{Name: Model: UUID: Version: Release: Serial: IsSanOptimized:false IsDisaggregated:false ZAPIsExist:true HasREST:false IsClustered:false}" remoteErr="connection error => Post "https://XX.XX.XX.XX:443/servlets/netapp.servlets.admin.XMLrequest_filer\": context deadline exceeded (Client.Timeout exceeded while awaiting headers)\nconnection error: Get "https://XX.XX.XX.XX/api/cluster?fields=*&return_records=true\": context deadline exceeded (Client.Timeout exceeded while awaiting headers)"

We cross-checked the config file for working environment and this, no differences are seen
Here no cluster information is being pulled and it even shows IsClustered as false which is incorrect. Also it shows HasREST as false
Config file has entries for both REST and ZAPI collectors. We tried mentioning REST in the top of list but no difference. We even tried with only REST and RESTPerf - result was same. REST service was disabled/enabled from cluster - same result
We also tried curl to the below location from systemshell of the cluster :

curl -k -u username:password https://XX.XX.XX.XX/api/cluster?fields=*&return_records=true

It was giving same result as the working environment when we compared with it
Also we added 60s timeout in the config file under Poller (currently nothing is mentioned) and found no changes in error.

Need some insights on what could be the issue here
Thank you.

[drowsy vessel]

hi @dusky niche you mentioned that you tried to curl from the systemshell of the cluster - have you also tried to curl /api/cluster from another machine? If that doesn't work, Harvest won't work either. Might be worth asking how to fix that in the #1062049169520476220 channel

[dusky niche]

Hello @drowsy vessel ,
I have referred the kb: https://kb.netapp.com/Cloud/Astra/Trident/How_to_test_connectivity_to_Trident_CSI_Controller_from_a_particular_Kubernetes_node

Tried doing a curl from the pod to the cluster and it was returning the details without any issues.
We did even verified the apache logs from cluster at the same tine and the responses are looking good. However when the pod is deleted and recreated and when harvest collects the metrics automatically, its failing.
HasREST is always showing false and "IsClustered" is false too, which is incorrect.
We tried mentioning only ZAPI and ZAPIPerf under pollers and tried, that time also same timeout issue -

context deadline exceeded (Client.Timeout exceeded while awaiting headers)\nconnection error:

Not sure if the issue is at Poller-end or something is incorrect in ONTAP end
We have also cross-checked the config files and all are matching with the working setup