Description

This article explains about log message: DRD0 XQSS_CMERROR_DSTAT_INT_REG_DROP0_QDEPTH_UNDRN appeared on FPC and solution to resolve it.

Symptoms

1) Interfaces on the affected FPC may experience flapping just before the occurrence of major errors, below logs will be reported.
 

mib2d[75768]: %DAEMON-4-SNMP_TRAP_LINK_DOWN: ifIndex 794, ifAdminStatus up(1), ifOperStatus down(2), ifName et-0/3/2

mib2d[75768]: %DAEMON-4-SNMP_TRAP_LINK_DOWN: ifIndex 587, ifAdminStatus up(1), ifOperStatus down(2), ifName et-0/3/3


 

2) Major error will appear on the FPC, leading to PFE shutdown and interface administration down as below. Services will be impacted as interface remains down.
 

Physical interface: et-1/3/3, Administratively down, Physical link is Down

 Interface index: 348, SNMP ifIndex: 794, Generation: 88012

 Link-level type: Ethernet, MTU: 9192, MRU: 9200, Speed: 100Gbps, BPDU Error: None, Loop Detect PDU Error: None, Loopback: Disabled, Source filtering: Disabled, Flow control: Enabled

 Pad to minimum frame size: Disabled

 Device flags  : Present Running Down PFE-Disabled

 Interface flags: Hardware-Down Down SNMP-Traps Internal: 0x4000

 Link flags    : None

 CoS queues    : 8 supported, 8 maximum usable queues

 Schedulers    : 0

 Hold-times    : Up 0 ms, Down 0 ms

 Damping       : half-life: 0 sec, max-suppress: 0 sec, reuse: 0, suppress: 0, state: unsuppressed

 Current address: 88:28:fb:b5:a5:8a, Hardware address: 88:28:fb:b5:a5:8a

 Last flapped  : 2023-11-19 10:53:52 IST (02:58:51 ago)

 Statistics last cleared: 2023-11-19 03:51:14 IST (10:01:29 ago)


 

3) The interface flapping failed to flush the queue, leading to the xqss error and triggering the underrun.
 

aftd-trio[16842]: %USER-3: [Error] JnhPlus: CosSchedNodeQueuesTrio: Failed to configure Queue nodes with baseQ:24 numQueues:^H pfeInst:4

ytchip-xqss[16840]: %USER-3: [Error] CDA: XQSS: Set Q node failed for qIndex:33


 

4) Below error log messages will appear on the FPC, which could result in major errors:
 

root[27513]: %USER-5: pfe_get_state.sh : PFE AFT Get state started - Module: (xqss), Error: (XQSS_CMERROR_DSTAT_INT_REG_DROP0_QDEPTH_UNDRN), Date/Time: (19/11/2023 05:23:40)

alarmd[75764]: %DAEMON-4: Alarm set: FPC id=150995048, color=RED, class=CHASSIS, reason=FPC 0 Major Errors
 

5) Device configured with high-priority scheduler:

# set class-of-service schedulers <scheduler name> priority high

Solution

Please refer below PR for more details and solution.

https://prsearch.juniper.net/problemreport/PR1638410

To recover instantly from this situation, please reboot the affected FPC but major errors may appear again if interfaces continuing to flap.

Modification History

2024-01-21 : Article Created