hpe-storage / array-exporter Goto Github PK

Exporter doesn't provide metrics about HW health, like hard drive health, HW components health, overall storage system health, temperatures, fans, etc.

Hi,

I need to modify the grafana dashboards to have a decent view.
All the datasources are configured to "null". They need to be replaced to "prometeus".

There's still a reference to "nimble" array in hpe-storage-volume-details_rev3.json
"definition": "label_values(hpenimble_pool_capacity_bytes, array)",

And the "pool" variables are statically configured. They can be set dynamicaly with a query like this one :
label_values(hpe_storagepool_capacity_bytes, storagepool)

Regards.

Gaël

I tried to configured the exporter to use the WSAPI enabled on Alletra MP, but it does not work. Is this model supported ? if yes, is there any configuration which needs to be performed ?

Hello,

I would like to know if it's possible to specify the API access port ?
It seems that we cannot change the port 5392.

I already tried address: x.x.x.x:8080, but it always append :5392 at the end.

Any idea or it's not possible ?

Thanks.

The config file storage-system.yaml

address: 10.10.10.1
username: exampleuser
password: examplepassword

We want use one array-exporter to monitor multiple HP3Par storage instance. The config is like:

address: 10.10.10.1
username: exampleuser
password: examplepassword
address: 10.10.10.2
username: exampleuser
password: examplepassword
...

Could you please support this feature?

The similiar repo is : https://github.com/Sispheor/hp3par-exporter

Exporter provide only one metric for free space of CPG. But web interface and commands like showspace provide different values, including deduplication and compression. But it's impossible right now to derive correct amount free space from available metrics.

HPE CSI Driver supports multi-tenancy. Nimble storage provider, in our use case, assigns folders per tenant to isolate and control access to pool capacity in the storage array. Volume and snapshot information is all available for the PVCs associated with the tenant controlled within the cluster.

A tenant is limited by protection templates, performance policies, usage limit, and overdraft limit. To effectively monitor, trend and alert when approaching limits, tenant specific metrics are needed.

Please consider expanding the metrics exposed by the Storage array exporter to include folder specific limits and usage.

It would be helpful to have some metrics which provide statistics / host, similar to statvlun -rw -d 5 -hostsum -iter 1 command . Only statistics for volumes do not help much when you have many volumes and not a specific naming convention. Not to mention when you have volumes mapped to ESX, it is difficult to drill down without statistics / host.
Also a metric similar to statpd for physical disks.