You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
We're finding that as we're adding more vcenters into zabbix it's getting harder to keep track of all the created items. Whilst the esx.discovery returns the cluster name so we can create a host prototype for the cluster HV's none of the others do ({#CLUSTER} is always datastore for the datastore discovery). If {#CLUSTER} or an equivalent was returned for VM's and datastores then we could use that for group prototypes making it significantly easier to keep track of the items.
The text was updated successfully, but these errors were encountered:
We're finding that as we're adding more vcenters into zabbix it's getting harder to keep track of all the created items. Whilst the esx.discovery returns the cluster name so we can create a host prototype for the cluster HV's none of the others do ({#CLUSTER} is always datastore for the datastore discovery). If {#CLUSTER} or an equivalent was returned for VM's and datastores then we could use that for group prototypes making it significantly easier to keep track of the items.
The text was updated successfully, but these errors were encountered: