Problem/Justification
Regression? Since SCALE, SNMP isn’t reporting zpool usage or zvol details anymore.
Impact
As these details were available in CORE and break existing monitoring configurations.
User Story
I upgraded from CORE to SCALE and miss useful monitoring metrics. Issue was reported to Jira by other users:
NAS-128121 also shows how to implement the zpool utilization object.
I am in the process of trying to update a copy the existing Zabbix monitoring template template_app_truenas_core_snmp.yaml to a corresponding scale version, and I would agree in calling removal of them a regression.
The missing truenas.zpool.size.total (.1.3.6.1.4.1.50536.1.1.1.1.4.{#SNMPINDEX}) and truenas.zpool.used (.1.3.6.1.4.1.50536.1.1.1.1.5.{#SNMPINDEX}) keys are really degrading the value of the template.
Funnily enough, the MIB that ought to expose/translate the OIDs is included in the /usr/local/share/snmp/mibs/ directory, as per the documentation here: SNMP | TrueNAS Documentation Hub
So, either they fix the missing SNMP extensions to report these values (quite obviously preferred), or they fix the documentation and say ‘we only do the barebones basics’ (which would make me a sad bunny).
