Multi-tenant reporting
We are planing to develop a number of checks for multi-tenant monitoring for NetApp, which is used to give internal or external clients (tenants) access to volumes or entire _Storage Virtual Machines _(SVMs).
Space and IOPS per tenant
The checks calculate the disk space and IOPS any given client uses on the system - i.e. the sum over all volumes that are assigned to this specific tenant.
- Disk space: calculates the sum of disk space used on all volumes assigned to the tenant.
- IOPS (Input Output Operations per Second): the IOPS for a tenant is equivalent to the sum of the IOPS on all volumes assigned to the tenant.
Assigning Volumes and SVMs to a tenant
How does the check identify volumes or SVMs that are assigned to any given tenant? Volumes and/or SVMs can be assigned to a client by using the name of the volume or SVM. This process is made a lot easier by using naming conventions, e.g. using the client’s initials at the end of an instance’s name, separated by an underscore. Furthermore it works dynamically, meaning there is no need for maintenance by the monitoring admin. In principle, the checks could also be given be a list of volumes and SVMs to be monitored. We might add this feature at a later stage.
Output of aggregated data
Output data will be returned as alarms according to the Nagios guidelines, performance data is outputted to STDOUT, with sum and values of the individual volumes. For alarms, sums are compared to threeholds, not the individual instance values. In addition we can provide interfaces according to customer request (SQL, CSV, …).
Examples
Please find configuration examples illustrating the check’s functionality in the description pages below:
Comments