snmpwalk returns 0 for Juniper MX ae interfaces

[mbuyukkarakas]

Dear all,

I'm trying to monitor a Junier MX device with SNMPv2.

For some of interfaces (named like : xe-x/x or et-x/x everything is ok. snmpwalk and zabbix are returning correct data.
But for the ae (aggregated) interfaces snmpwalk or snmpget are always returning the bits send/received data as 0. But I'm sure traffic flows over that interfaces.
Customer says they are able to monitor data flow with PRTG.

Anybody ever seen something like this ? Can you help me please ?
Thank you.

# snmpget -v 2c -c xxx 10.40.16.11 .1.3.6.1.2.1.31.1.1.1.6.630
IF-MIB::ifHCInOctets.630 = Counter64: 0

# snmpwalk -v 2c -c xxx 10.40.16.11 .1.3.6.1.2.1.31.1.1.1.6.630
IF-MIB::ifHCInOctets.630 = Counter64: 0

[tim.mooney]

Perhaps ask the customer if they know whether PRTG is also using SNMP (and if so, what version), or whether it's getting the data via some other method (perhaps JunOS commands?)

Zabbix's SNMP monitors use the underlying Linux net-snmp libraries, so it's no surprise that Zabbix and snmpwalk get the same results.

Assuming the customer has Juniper support, they may be able to ask Juniper support what needs to be done to get correct data via SNMP for the aggregated interfaces. It might be a firmware or JunOS issue, or it could be something else.

[semaja2]

Did you ever find a fix to this? We are having the same issue and google searches are not showing any helpful results yet

For our situation it seems to occasionally return 0, which causes the metrics to go out of whack