#[NABOX] Cluster missing in Grafana

1 messages · Page 1 of 1 (latest)

thorn herald
#

Hi All,
I've just upgraded to NABox 4.0.3. Migrated all Data succesfully.
But there are some difference between the new Nabox vs. the old v3.5.3.

On both instances the admin page report 2 out of 3 systems reporting actively. Which for us is to expected. the first two systems is a 4-node MCC-IP cluster and the other is merely a Snapmirror DR target, not servicing data at the front end ports.
The difference lies in Grafana. Where we used to have all systems visible for certain metrics (ex. cdot Node Details, cdot Cluster). On the new instance only the two MCC-IP clusters are visible. The whole snapmirror DR target is not there. Am I missing something.

My main question is it me or is this how the new NABox 4 is setup? Like as long as there is no data served, no metric are avaible in Grafana.
See the screenshots for more info. One of the two DC's and one with only one DC.

In addition, although the snapmirror DR is not doing host IO we would still like to see the node load etc.

amber light
#

Maybe a difference in how harvest gets the variables ? Are you sure data is actually collecting ?

thorn herald
#

In the old Nabox 3.5.3 it is collecting something at least (see screenshot of the last 15 mins). And it is using Harvest 24.05.1-1. One version behind the current one. Let me see if I create a test svm and transfer some data. It should be visible in grafana on the new host.

ruby robin
thorn herald
#

Hi Chris, I cannot send the logs due the size of the attachment exceding 20MB. Can I upload some where?

thorn herald
#

I've uploaded the support bundle.

#

nabox-logs-2024-07-09_130912.tgz

#

Thank you!

ruby robin
thorn herald
#

I will look into that. All nodes shuld have the same permissions. But I will check to be sure.

ruby robin
#

I can share the poller name with the permissions errors if that is helpful. It is a single poller. The other pollers are fine

thorn herald
#

i've changed the user role from a specially made harvest role to full readonly. But still I get this. init collector-object error="StatusCode: 401, Error: auth failed, Message: 401 Unauthorized, API: /api/cluster?fields=%2A&return_records=true". So I need further investigation on the netapp side.

odd gazelle
#

Did u restart Harvest after making permission changes for User in ONTAP?

thorn herald
#

uhmmm. No? 😬 . Lemme try that first

odd gazelle
#

Yes pls.