Cisco N9K-C9396PX with N2K-C2232PP-10GE missing fex interface data

Than this is only device/firmware issue. Try update firmware

thank you, for helping, setting max rep to 0 didn't help.

Note, disable max max rep for device can significant increase polling time of device.

Seems your device have troubles with SNMP max rep values. It's also firmware issue.

You can try reduce snmp max value (to 10 or 20) or complete disable for device (set 0).
In device edit snmp tab set the appropriate value, wait for 10 minus (for 2 poller rounds) and check if graphs return.

Also check if you use latest firmware for your platform, probably issue already fixed.
Latest for your platform is 9.3(11), but you use 9.3(9).

Yes, but the device doesn't return this data when this query is run: 

/usr/bin/snmpbulkwalk -Cr'50' -t '120' -r '5' -v3 -l 'authPriv' -n '' -x *** -X *** -a *** -A *** -u *** -Pud -OQUs -m IF-MIB -M /opt/observium/mibs/rfc:/opt/observium/mibs/net-snmp 'udp':'nexus1.wavecom.ee':'161' ifXEntry

I've also added a screenshot showing that querying directly I get a result.

Hello

Using mib browser to query this interface, I seem to get responses to the OIDs that are missing in poller.output for 530841856 interface.
Please see the attached screenshot

This device is not returning counters for these ports. This is probably something you need to ask Cisco about.

The port in the screenshot is ifIndex 530841856

You can see the lack of counters :

root@web:~# grep '165/1/5' poller.output
ifDescr.530841856 = Ethernet165/1/5
ifName.530841856 = Ethernet165/1/5
| 530841856  | Et165/1/5  | Ethernet                    | 10Gbps  | 0bps     | 0bps     | 0B      | 0B       | 0      | 0       | Y  |
root@web:~#

root@web:~# grep '530841856' poller.output | grep ^if
ifIndex.530841856 = 530841856
ifDescr.530841856 = Ethernet165/1/5
ifType.530841856 = ethernetCsmacd
ifMtu.530841856 = 9216
ifSpeed.530841856 = 4294967295
ifPhysAddress.530841856 = 58:a:20:37:8a:c6
ifAdminStatus.530841856 = up
ifOperStatus.530841856 = up
ifLastChange.530841856 = 0:0:00:00.00
ifOutQLen.530841856 = 0
ifSpecific.530841856 = zeroDotZero
ifName.530841856 = Ethernet165/1/5
ifLinkUpDownTrapEnable.530841856 = enabled
ifHighSpeed.530841856 = 10000
ifPromiscuousMode.530841856 = false
ifConnectorPresent.530841856 = true
ifAlias.530841856 = Core: NAS-10G
root@web:~#

Where as the slot 1 ports seem to have counters:

root@web:~# grep '1241513984' poller.output | grep ^if
ifIndex.1241513984 = 1241513984
ifDescr.1241513984 = ii1/1/1
ifType.1241513984 = ethernetCsmacd
ifMtu.1241513984 = 9216
ifSpeed.1241513984 = 4294967295
ifPhysAddress.1241513984 =
ifAdminStatus.1241513984 = up
ifOperStatus.1241513984 = up
ifLastChange.1241513984 = 0:0:00:00.00
ifInOctets.1241513984 = 3868081343
ifInUcastPkts.1241513984 = 633757751
ifInNUcastPkts.1241513984 = 470411029
ifInDiscards.1241513984 = 0
ifInErrors.1241513984 = 0
ifInUnknownProtos.1241513984 = 0
ifOutOctets.1241513984 = 2013542020
ifOutUcastPkts.1241513984 = 2405346490
ifOutNUcastPkts.1241513984 = 2249443135
ifOutDiscards.1241513984 = 482
ifOutErrors.1241513984 = 0
ifOutQLen.1241513984 = 0
ifSpecific.1241513984 = zeroDotZero
ifName.1241513984 = ii1/1/1
ifInMulticastPkts.1241513984 = 453422264
ifInBroadcastPkts.1241513984 = 16988765
ifOutMulticastPkts.1241513984 = 1679766167
ifOutBroadcastPkts.1241513984 = 569676968
ifHCInOctets.1241513984 = 1048761867320511
ifHCInUcastPkts.1241513984 = 1078670549047
ifHCInMulticastPkts.1241513984 = 453422264
ifHCInBroadcastPkts.1241513984 = 16988765
ifHCOutOctets.1241513984 = 537551530374788
ifHCOutUcastPkts.1241513984 = 526391356602
ifHCOutMulticastPkts.1241513984 = 14564668055
ifHCOutBroadcastPkts.1241513984 = 4864644264
ifLinkUpDownTrapEnable.1241513984 = disabled
ifHighSpeed.1241513984 = 42000
ifPromiscuousMode.1241513984 = false
ifConnectorPresent.1241513984 = false
ifAlias.1241513984 =
ifCounterDiscontinuityTime.1241513984 = 284:2:23:54.47
root@web:~#

We collect port traffic counters from IF-MIB, probably the most standardised MIB that exists. There's almost zero scope for any thing on our side to affect collecting this that wouldn't break everything.

This is a Cisco bug.

Thanks,

adam.

sorry, I've attached the output as poller.output

Oh, I mean please attach this debug output to issue tracker.