Wrong RPDU SNMP reply

[Anonymous user]

Dear all,

Please advice whats wrong:

AP8881 PDU connected with 3 other via NPS

When I do SNMP get for power counters the wrong PDU replies:

[root@cacti ~]# snmpget -c public -v1 -On 172.27.4.3 .1.3.6.1.4.1.318.1.1.26.4.3.1.5.1
.1.3.6.1.4.1.318.1.1.26.4.3.1.5.1 = INTEGER: 112
[root@cacti ~]# snmpget -c public -v1 -On 172.27.4.3 .1.3.6.1.4.1.318.1.1.26.4.3.1.5.2
.1.3.6.1.4.1.318.1.1.26.4.3.1.5.2 = INTEGER: 15
[root@cacti ~]# snmpget -c public -v1 -On 172.27.4.3 .1.3.6.1.4.1.318.1.1.26.4.3.1.5.3
.1.3.6.1.4.1.318.1.1.26.4.3.1.5.3 = INTEGER: 14
[root@cacti ~]# snmpget -c public -v1 -On 172.27.4.3 .1.3.6.1.4.1.318.1.1.26.4.3.1.5.4
.1.3.6.1.4.1.318.1.1.26.4.3.1.5.4 = INTEGER: 120

RPDU group screenshot:

Did APC changed something in OIDs?

[voidstar_apc]

I'm not familiar with the RPDUs but I took a look at the MIB and noticed the table contains OIDs that describe which RPDU each table entry corresponds with.

You're currently looking at:

.1.3.6.1.4.1.318.1.1.26.4.3.1.5

You might consider also looking at:

.1.3.6.1.4.1.318.1.1.26.4.3.1.2

and

.1.3.6.1.4.1.318.1.1.26.4.3.1.3

to determine the corresponding RPDU module for the table index.

[BillP]

Agree with voidstar. The SNMP indexes may not be the same as the display ID since the user can change the display ID after power up. As voidstar mentioned, you should check the index table with the specific OID mentioned. If it is confusing, you'd have to reset the indexes or you can change the module IDs around to match the index. I believe to reset the indexes it could be painful and require a reset to defaults and playing around with the group communication (removing PDUs and re-adding them, etc).

There is a bit of a note on that topic here -> http://www.apc.com/us/en/faqs/FA164436

Network Port Sharing and SNMP Polling

All Rack PDUs in a group can be accessed through the Host Rack PDU via SNMP "rPDU2" OIDs available in our PowerNet MIB.

Note: PowerNet MIB v4.0.6 or higher is required for use with SNMP and a Network Port Sharing enabled device.

The full path to these OIDs is:
iso(1).org(3).dod(6).internet(1).private(4).enterprises(1).apc(318).products(1).hardware(1).rPDU2(26)

Individual Rack PDUs can be identified in the SNMP MIB tables by viewing the corresponding "Module" OIDs in each table. These Module OIDs will return the Display ID of the Rack PDU.

Example Module OIDs: rPDU2IdentModule, rPDU2DeviceConfigModule, rPDU2SensorTempHumidityConfigModule

In order to be backwards compatible with previous versions, the Host Rack PDU will always be the first index in any table that supports multiple Rack PDUs. In addition, after the Rack PDU group is set up, the index order of guest Rack PDUs should not change even if the Display ID is changed or a PDU temporarily loses communication. The index order should only change if you manually remove a Rack PDU from the group. A MIB table walk should skip the indexes associated with a Rack PDU that has temporarily lost communication.

[Anonymous user]

As I understand there is no any OID to get SNMP power counters by DisplayID?

[voidstar_apc]

There is by way of a relational join, you need to get:

.1.3.6.1.4.1.318.1.1.26.4.3.1.5.x

where .1.3.6.1.4.1.318.1.1.26.4.3.1.2.x == DisplayID

[Anonymous user · Accepted solution]

thanks for reply but that is not comfortable way. If someone reorder DisplayID than need to change OIDs in all monitoring systems.

Better not touch Display IDs since PDU group powered on.