2015-08-10 01:15 PM
I am building an internal reporting solution for the performance metrics of Clustered ONTAP.
Basically, i am using OnCommand Performance Manager (version 1.1) as Data Collector, InfluxDB as back-end (instead of Graphite) and Grafana for the dashboard aspects.
Many of the important metrics are well-forwarded from OPM to InfluxDB but i don't see any informations regarding throughput at volume and LUN level.
NetApp KB (https://kb.netapp.com/support/index?page=content&id=1014930&locale=en_US) only speaks about READ/WRITE OPS or latency.
Is there an easy way to retrieve there metrics?
2015-08-12 12:59 PM
OPM has a fixed counter catalog and luns have READ_DATA and WRITE_DATA but volumes do not, which surprised me. If you need these you could use harvest, an alternative community supported data collector which forwards metrics to Graphite/InfluxDB which will be posting to the toolchest shortly.
Storage Architect, NetApp EMEA
2016-01-12 08:32 AM
Do you know when the community supported data collector which forwards metrics to Graphite/InfluxDB will be posted to the toolchest?
2016-01-13 12:21 AM
I'm using an existing InfluxDB instance and the dashboards in Grafana dashboard don't show any data. I can see data going into InfluxDB from the Filers but think the Dashboards may need updating to use point at that data?
2016-01-13 12:36 AM
Yes, the default Grafana dashboards provided by Harvest are defined with Graphite metric strings and functions. So while Harvest can send data to InfluxDB since it accepts the Graphite newline seperated metrics format you will not be able to use the default dashboards as-is but rather will need to refactor to work with InfluxDB. Sorry I don't have better news for you.
Storage Architect, NetApp EMEA (and author of Harvest)
Blog: It all begins with data
2016-01-13 05:30 AM
No problem thanks for the quick response. We're running up a graphite instance to see if we can map the data across to InfluxDB to alter the dashboards.