node_new_status false positives? | NetApp | Page 1

rancid cosmos Sep 12, 2023, 5:06 PM

#

On 23.05 and the clusters are 9.9.1P15

Has anyone else had issues with node_new_status randomly dropping to 0 for blips? This is happening randomly across multiple clusters, with so far no apparent cause on the ONTAP side, and it's causing false positive headaches for the customer.

This is possibly a networking problem, and I'm waiting on the /var/log/harvest/ files to get more insight.

I am also currently investigating more reliable ways for them to alert on a node being down. Perhaps that is not the best one to use in the first place.

rancid cosmos Sep 12, 2023, 5:42 PM

#

After looking at this more, seems like this is more of a metric for node health and not for the node being up or down #doh

#

https://giphy.com/gifs/MicrosoftCloud-clippy-msbuild-microsoft-build-jR8yFMATIGVAY9mTZu

Giphy

Microsoft Cloud

loud orchid Sep 13, 2023, 9:43 AM

#

@rancid cosmos , we are fetching node health status here https://github.com/NetApp/harvest/blob/main/conf/zapi/cdot/9.8.0/node.yaml#L20 and then converting to 0/1 by value_to_num. Yes, you are right, it's for healthyness status of node and not for node up/down info. Screenshot have the detail of system-node-get-iter zapi response

#node_new_status false positives?