Introdução

Este documento descreve como identificar o computador e o switch leaf para uma plataforma de implantação nativa de nuvem (CNDP) específica da Session Management Function (SMF) e resolver o alerta "network-receive-error" relatado no Common Execution Environment (CEE).

Problema

Os alertas de "erro de recepção de rede" são reportados no CEE Opcenter Rack2.

[lab0200-smf/labceed22] cee# show alerts active summary
NAME UID SEVERITY STARTS AT SOURCE SUMMARY
-----------------------------------------------------------------------------------------------------------------------------------------------------
network-receive-error 998c77d6a6a0 major 10-26T00:10:31 lab0200-smf-mas Network interface "bd0" showing receive errors on hostname lab0200-s...
network-receive-error ea4217bf9d9e major 10-26T00:10:31 lab0200-smf-mas Network interface "bd0" showing receive errors on hostname lab0200-s...
network-receive-error 97fad40d2a58 major 10-26T00:10:31 lab0200-smf-mas Network interface "eno6" showing receive errors on hostname lab0200-...
network-receive-error b79540eb4e78 major 10-26T00:10:31 lab0200-smf-mas Network interface "eno6" showing receive errors on hostname lab0200-...
network-receive-error e3d163ff4012 major 10-26T00:10:01 lab0200-smf-mas Network interface "bd0" showing receive errors on hostname lab0200-s...
network-receive-error 12a7b5a5c5d5 major 10-26T00:10:01 lab0200-smf-mas Network interface "eno6" showing receive errors on hostname lab0200-...

Consulte o Ultra Cloud Core Subscriber Microservices Infrastructure Operations Guide para obter a descrição do alerta. 

Alert: network-receive-errors
 Annotations:
  Type: Communications Alarm
  Summary: Network interface "{{ $labels.device }}" showing receive errors on hostname {{ $labels.hostname }}"
  Expression:
   |
   rate(node_network_receive_errs_total{device!~"veth.+"}[2m]) > 0
For: 2m
Labels:
 Severity: major

Identificar a Origem dos Alertas

Faça login no CEE labceed22, verifique os detalhes do alerta "network-receive-error" reportados nas interfaces bd0 e eno6 para identificar o nó e o pod.

[lab0200-smf/labceed22] cee# show alerts active summary
NAME                   UID           SEVERITY  STARTS AT       SOURCE                 SUMMARY                                                            
---------------------------------------------------------------------------------------------------------------------------------------------------------
network-receive-error  3b6a0a7ce1a8  major     10-26T21:17:01  lab0200-smf-mas  Network interface "bd0" showing receive errors on hostname tpc...  
network-receive-error  15abab75c8fc  major     10-26T21:17:01  lab0200-smf-mas  Network interface "eno6" showing receive errors on hostname tp...  

Execute show alerts ative detail network-receive-error <UID> para obter detalhes do alerta.

No exemplo, a origem de ambos os alertas é o nó lab0200-smf-primary-1 pod node-export-47xmm.

[lab0200-smf/labceed22] cee# show alerts active detail network-receive-error  3b6a0a7ce1a8
alerts active detail network-receive-error 3b6a0a7ce1a8
 severity    major
 type        "Communications Alarm"
 startsAt    2021-10-26T21:17:01.913Z
 source      lab0200-smf-primary-1
 summary     "Network interface \"bd0\" showing receive errors on hostname lab0200-smf-primary-1\""
 labels      [ "alertname: network-receive-errors" "cluster: lab0200-smf_cee-labceed22" "component: node-exporter" "controller_revision_hash: 75c4cb979f" "device: bd0" "hostname: lab0200-smf-primary-1" "instance: 10.192.1.42:9100" "job: kubernetes-pods" "monitor: prometheus" "namespace: cee-labceed22" "pod: node-exporter-47xmm" "pod_template_generation: 1" "replica: lab0200-smf_cee-labceed22" "severity: major" ]
 annotations [ "summary: Network interface \"bd0\" showing receive errors on hostname lab0200-smf-primary-1\"" "type: Communications Alarm" ]

[lab0200-smf/labceed22] cee# show alerts active detail network-receive-error 15abab75c8fc
alerts active detail network-receive-error 15abab75c8fc
 severity    major
 type        "Communications Alarm"
 startsAt    2021-10-26T21:17:01.913Z
 source      lab0200-smf-primary-1
 summary     "Network interface \"eno6\" showing receive errors on hostname lab0200-smf-primary-1\""
 labels      [ "alertname: network-receive-errors" "cluster: lab0200-smf_cee-labceed22" "component: node-exporter" "controller_revision_hash: 75c4cb979f" "device: eno6" "hostname: lab0200-smf-primary-1" "instance: 10.192.1.42:9100" "job: kubernetes-pods" "monitor: prometheus" "namespace: cee-labceed22" "pod: node-exporter-47xmm" "pod_template_generation: 1" "replica: lab0200-smf_cee-labceed22" "severity: major" ]
 annotations [ "summary: Network interface \"eno6\" showing receive errors on hostname lab0200-smf-primary-1\"" "type: Communications Alarm" ]

Validar Status de Nó, Pod e Portas

Validação de nó e pod do VIP principal

Faça login no VIP primário do K8s do Rack2 para validar o status do nó e do pod de origem.

No exemplo, ambos estão em um bom estado: Pronto e em execução.

cloud-user@lab0200-smf-primary-1:~$ kubectl get nodes
NAME                         STATUS   ROLES                  AGE    VERSION
lab0200-smf-primary-1   Ready    control-plane          105d   v1.21.0
lab0200-smf-primary-2   Ready    control-plane          105d   v1.21.0
lab0200-smf-primary-3   Ready    control-plane          105d   v1.21.0
lab0200-smf-worker-1   Ready    <none>                 105d   v1.21.0
lab0200-smf-worker-2   Ready    <none>                 105d   v1.21.0
lab0200-smf-worker-3   Ready    <none>                 105d   v1.21.0
lab0200-smf-worker-4   Ready    <none>                 105d   v1.21.0
lab0200-smf-worker-5   Ready    <none>                 105d   v1.21.0

cloud-user@lab0200-smf-primary-1:~$ kubectl get pods -A -o wide | grep node-exporter--47xmm
cee-labceed22     node-exporter-47xmm                                       1/1     Running   0          18d    10.192.1.44       lab0200-smf-primary-1   <none>           <none>

Validações de portas do VIP principal do K8s

Valide se as interfaces bd0 e eno6 estão ATIVADAS com endereço IP | grep eno6 e ip addr | grep bd0.

Note: Quando o filtro é aplicado para bd0, o eno6 é mostrado na saída. O motivo é que o eno5 e o eno6 são configurados como interfaces vinculadas em bd0, que podem ser validadas no SMI Cluster Deployer.

cloud-user@lab0200-smf-primary-1:~$ ip addr | grep eno6
3: eno6: <BROADCAST,MULTICAST,SECONDARY,UP,LOWER_UP> mtu 1500 qdisc mq primary bd0 state UP group default qlen 1000

cloud-user@lab0200-smf-primary-1:~$ ip addr | grep bd0
2: eno5: <BROADCAST,MULTICAST,SECONDARY,UP,LOWER_UP> mtu 1500 qdisc mq primary bd0 state UP group default qlen 1000
3: eno6: <BROADCAST,MULTICAST,SECONDARY,UP,LOWER_UP> mtu 1500 qdisc mq primary bd0 state UP group default qlen 1000
12: bd0: <BROADCAST,MULTICAST,PRIMARY,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group default qlen 1000
13: vlan111@bd0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group default qlen 1000
14: vlan112@bd0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group default qlen 1000
182: cali7a166bd093d@if4: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1440 qdisc noqueue state UP group default

Validações de Portas do Implantador de Cluster SMI

Faça login no Cluster Manager VIP e, em seguida, acesse o ssh para Operations (Ops) Center ops-center-smi-cluster-deployer.

cloud-user@lab-deployer-cm-primary:~$ kubectl get svc -n smi-cm
NAME                                          TYPE        CLUSTER-IP       EXTERNAL-IP     PORT(S)                                                 AGE
cluster-files-offline-smi-cluster-deployer    ClusterIP   10.102.53.184    <none>          8080/TCP                                                110d
iso-host-cluster-files-smi-cluster-deployer   ClusterIP   10.102.38.70     172.16.1.102   80/TCP                                                  110d
iso-host-ops-center-smi-cluster-deployer      ClusterIP   10.102.83.54     172.16.1.102   3001/TCP                                                110d
netconf-ops-center-smi-cluster-deployer       ClusterIP   10.102.196.125   10.241.206.65   3022/TCP,22/TCP                                         110d
ops-center-smi-cluster-deployer               ClusterIP   10.102.12.170    <none>          8008/TCP,2024/TCP,2022/TCP,7681/TCP,3000/TCP,3001/TCP   110d
squid-proxy-node-port                         NodePort    10.102.72.168    <none>          3128:32572/TCP                                          110d

cloud-user@lab-deployer-cm-primary:~$ ssh -p 2024 admin@10.102.12.170
admin@10.102.12.170's password:
      Welcome to the Cisco SMI Cluster Deployer on lab-deployer-cm-primary
      Copyright © 2016-2020, Cisco Systems, Inc.
      All rights reserved.
admin connected from 172.16.1.100 using ssh on ops-center-smi-cluster-deployer-5cdc5f94db-bnxqt
[lab-deployer-cm-primary] SMI Cluster Deployer# 

Verifique o cluster, os padrões de nó, as interfaces e o modo de parâmetros do nó. No exemplo, o lab0200-smf.

[lab-deployer-cm-primary] SMI Cluster Deployer#  show running-config clusters
clusters lab0200-smf
 environment lab0200-smf-deployer_1
…
 node-defaults initial-boot netplan ethernets eno5
  dhcp4 false
  dhcp6 false
 exit
 node-defaults initial-boot netplan ethernets eno6
  dhcp4 false
  dhcp6 false
 exit
 node-defaults initial-boot netplan ethernets enp216s0f0
  dhcp4 false
  dhcp6 false
 exit
 node-defaults initial-boot netplan ethernets enp216s0f1
  dhcp4 false
  dhcp6 false
 exit
 node-defaults initial-boot netplan ethernets enp94s0f0
  dhcp4 false
  dhcp6 false
 exit
 node-defaults initial-boot netplan ethernets enp94s0f1
  dhcp4 false
  dhcp6 false
 exit
node-defaults initial-boot netplan bonds bd0
  dhcp4      false
  dhcp6      false
  optional   true
  interfaces [ eno5 eno6 ]
  parameters mode      active-backup
  parameters mii-monitor-interval 100
  parameters fail-over-mac-policy active
 exit

No VIP primário, valide erros e/ou quedas nas interfaces bd0 e eno6.

Quando ambas as interfaces têm quedas, o hardware do switch UCS ou Leaf precisa ser verificado para verificar se há problemas de hardware.

cloud-user@lab0200-smf-primary-1:~$  ifconfig bd0
bd0: flags=5187<UP,BROADCAST,RUNNING,PRIMARY,MULTICAST>  mtu 1500
        inet6 fe80::8e94:1fff:fef6:53cd  prefixlen 64  scopeid 0x20<link>
        ether 8c:94:1f:f6:53:cd  txqueuelen 1000  (Ethernet)
        RX packets 47035763777  bytes 19038286946282 (19.0 TB)
        RX errors 49541  dropped 845484  overruns 0  frame 49541
        TX packets 53797663096  bytes 32320571418654 (32.3 TB)
        TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

cloud-user@lab0200-smf-primary-1:~$  ifconfig eno6
eno6: flags=6211<UP,BROADCAST,RUNNING,SECONDARY,MULTICAST>  mtu 1500
        ether 8c:94:1f:f6:53:cd  txqueuelen 1000  (Ethernet)
        RX packets 47035402290  bytes 19038274391478 (19.0 TB)
        RX errors 49541  dropped 845484  overruns 0  frame 49541
        TX packets 53797735337  bytes 32320609021235 (32.3 TB)
        TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

Identificar o servidor UCS

Validação do servidor UCS do implantador de cluster SMI 

Execute show running-config clusters <nome do cluster> nodes <nome do nó> no SMI Cluster Deployer para descobrir o endereço IP do CIMC do servidor UCS.

[lab-deployer-cm-primary] SMI Cluster Deployer# show running-config clusters lab0200-smf nodes primary-1
clusters lab0200-smf
nodes primary-1
  maintenance  false
  host-profile cp-data-r2-sysctl
  k8s node-type       primary
  k8s ssh-ip          10.192.1.42
  k8s sshd-bind-to-ssh-ip true
  k8s node-ip         10.192.1.42
  k8s node-labels smi.cisco.com/node-type oam
  exit
  k8s node-labels smi.cisco.com/node-type-1 proto
  exit
  ucs-server cimc user admin
...
  ucs-server cimc ip-address 172.16.1.62
...
  exit

Use SSH para acessar o endereço IP do CIMC 172.16.1.62 através do CM ativo e valide o nome do servidor.

No exemplo, o nome do servidor é LAB0200-Server8-02.

cloud-user@lab-deployer-cm-primary:~$ ssh admin@172.16.1.62
Warning: Permanently added '172.16.1.62' (RSA) to the list of known hosts.
admin@172.16.1.62's password:
LAB0200-Server8-02#

Note: Valide o nome do servidor no CIQ (Questionário de informações do cliente), se o CIQ estiver disponível.

Mapeie as principais portas VIP e interfaces de rede UCS

No VIP primário, verifique os nomes de interface física para o eno6 com o comando ls -la /sys/class/net. No exemplo, quando o lscpi é usado para identificar o dispositivo eno6, a porta 1d:00.1 deve ser usada para identificar o eno6.

cloud-user@lab0200-smf-primary-1:~$ ls -la /sys/class/net
total 0
drwxr-xr-x  2 root root    0 Oct 12 06:18 .
drwxr-xr-x 87 root root    0 Oct 12 06:18 ..
lrwxrwxrwx  1 root root    0 Oct 12 06:18 bd0 -> ../../devices/virtual/net/bd0
lrwxrwxrwx  1 root root    0 Oct 12 06:18 bd1 -> ../../devices/virtual/net/bd1
…
lrwxrwxrwx  1 root root    0 Oct 12 06:18 eno5 -> ../../devices/pci0000:17/0000:17:00.0/0000:18:00.0/0000:19:01.0/0000:1b:00.0/0000:1c:00.0/0000:1d:00.0/net/eno5
lrwxrwxrwx  1 root root    0 Oct 12 06:18 eno6 -> ../../devices/pci0000:17/0000:17:00.0/0000:18:00.0/0000:19:01.0/0000:1b:00.0/0000:1c:00.0/0000:1d:00.1/net/eno6

Note: O lspci mostra informações sobre todos os dispositivos no servidor UCS, como MLOM, SLOM, PCI, etc. As informações do dispositivo podem ser usadas para mapear com os nomes das interfaces na saída do comando ls -la /sys/class/net.

No exemplo, a porta 1d:00.1 pertence à interface MLOM e eno6. O eno5 é uma porta 1d:00.0 MLOM.

cloud-user@lab0200-smf-primary-1:~$ lspci
……
1d:00.0 Ethernet controller: Cisco Systems Inc VIC Ethernet NIC (rev a2)
1d:00.1 Ethernet controller: Cisco Systems Inc VIC Ethernet NIC (rev a2)
3b:00.0 Ethernet controller: Intel Corporation Ethernet Controller 10G X550T (rev 01)
3b:00.1 Ethernet controller: Intel Corporation Ethernet Controller 10G X550T (rev 01)
5e:00.0 Ethernet controller: Intel Corporation Ethernet Controller XL710 for 40GbE QSFP+ (rev 02)
5e:00.1 Ethernet controller: Intel Corporation Ethernet Controller XL710 for 40GbE QSFP+ (rev 02)
d8:00.0 Ethernet controller: Intel Corporation Ethernet Controller XL710 for 40GbE QSFP+ (rev 02)
d8:00.1 Ethernet controller: Intel Corporation Ethernet Controller XL710 for 40GbE QSFP+ (rev 02)

Na GUI do CIMC, corresponda ao endereço MAC MLOM visto na saída de ifconfig do VIP principal.

cloud-user@lab0200-smf-primary-1:~$  ifconfig bd0
bd0: flags=5187<UP,BROADCAST,RUNNING,PRIMARY,MULTICAST>  mtu 1500
        inet6 fe80::8e94:1fff:fef6:53cd  prefixlen 64  scopeid 0x20<link>
        ether 8c:94:1f:f6:53:cd  txqueuelen 1000  (Ethernet)
        RX packets 47035763777  bytes 19038286946282 (19.0 TB)
        RX errors 49541  dropped 845484  overruns 0  frame 49541
        TX packets 53797663096  bytes 32320571418654 (32.3 TB)
        TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0


cloud-user@lab0200-smf-primary-1:~$  ifconfig eno6
eno6: flags=6211<UP,BROADCAST,RUNNING,SECONDARY,MULTICAST>  mtu 1500
        ether 8c:94:1f:f6:53:cd  txqueuelen 1000  (Ethernet)
        RX packets 47035402290  bytes 19038274391478 (19.0 TB)
        RX errors 49541  dropped 845484  overruns 0  frame 49541
        TX packets 53797735337  bytes 32320609021235 (32.3 TB)
        TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

Identificar o Switch Leaf

Na rede do Cluster Manager, como mostrado na imagem, o MLOM (eno5/eno6) está conectado aos Leafs 1 e 2.

Note: Validar deixa os nomes de host no CIQ, se o CIQ estiver disponível.

Efetue login em ambos os Leaves e gere o nome do servidor. 

No exemplo, as interfaces LAB0200-Server8-02 MLOM e MLOM estão conectadas às interfaces Eth1/49 em Leaf1 e Leaf2.

Leaf1# sh int description | inc LAB0200-Server8-02
Eth1/10       eth    40G     PCIE-01-2-LAB0200-Server8-02
Eth1/30       eth    40G     PCIE-02-2-LAB0200-Server8-02
Eth1/49       eth    40G     LAB0200-Server8-02 MLOM-P2


Leaf2# sh int description | inc LAB0200-Server8-02
Eth1/10       eth    40G     PCIE-01-1-LAB0200-Server8-02
Eth1/30       eth    40G     PCIE-02-1-LAB0200-Server8-02
Eth1/49       eth    40G     LAB0200-Server8-02 MLOM-P1

Solução

Importante:  Cada questão precisa de uma análise própria. Caso nenhum erro seja encontrado no lado do Nexus, verifique se há erros nas interfaces do servidor UCS.

No cenário, o problema está relacionado à falha de link na Leaf1 int eth1/49 que está conectada ao LAB0200-Server8-02 MLOM eno6.

O servidor UCS foi validado e nenhum problema de hardware foi encontrado, o MLOM e as portas estavam em bom estado.

A folha1 mostrou erros de saída TX:

Leaf1# sh int Eth1/49
Ethernet1/49 is up
admin state is up, Dedicated Interface
  Hardware: 10000/40000/100000 Ethernet, address: e8eb.3437.48ca (bia e8eb.3437.48ca)
  Description: LAB0200-Server8-02 MLOM-P2
  MTU 9216 bytes, BW 40000000 Kbit , DLY 10 usec
  reliability 255/255, txload 1/255, rxload 1/255
  Encapsulation ARPA, medium is broadcast
  Port mode is trunk
  full-duplex, 40 Gb/s, media type is 40G
  Beacon is turned off
  Auto-Negotiation is turned on  FEC mode is Auto
  Input flow-control is off, output flow-control is off
  Auto-mdix is turned off
  Rate mode is dedicated
  Switchport monitor is off
  EtherType is 0x8100
  EEE (efficient-ethernet) : n/a
    admin fec state is auto, oper fec state is off
  Last link flapped 5week(s) 6day(s)
  Last clearing of "show interface" counters never
  12 interface resets
  Load-Interval #1: 30 seconds
    30 seconds input rate 162942488 bits/sec, 26648 packets/sec
    30 seconds output rate 35757024 bits/sec, 16477 packets/sec
    input rate 162.94 Mbps, 26.65 Kpps; output rate 35.76 Mbps, 16.48 Kpps
  Load-Interval #2: 5 minute (300 seconds)
    300 seconds input rate 120872496 bits/sec, 22926 packets/sec
    300 seconds output rate 54245920 bits/sec, 17880 packets/sec
    input rate 120.87 Mbps, 22.93 Kpps; output rate 54.24 Mbps, 17.88 Kpps
  RX
    85973263325 unicast packets  6318912 multicast packets  55152 broadcast packets
    85979637389 input packets  50020924423841 bytes
    230406880 jumbo packets  0 storm suppression bytes
    0 runts  0 giants  0 CRC  0 no buffer
    0 input error  0 short frame  0 overrun   0 underrun  0 ignored
    0 watchdog  0 bad etype drop  0 bad proto drop  0 if down drop
    0 input with dribble  0 input discard
    0 Rx pause
  TX
    76542979816 unicast packets  88726302 multicast packets  789768 broadcast packets
    76632574981 output packets  29932747104403 bytes
    3089287610 jumbo packets
    79095 output error  0 collision  0 deferred  0 late collision
    0 lost carrier  0 no carrier  0 babble  0 output discard
    0 Tx pause

O alerta "network-receive-error" foi resolvido com a substituição de cabo no int eth1/49 Leaf1.

A última falha de link de interface foi relatada logo antes da substituição do cabo.

2021 Nov 17 07:36:48 TPLF0201 %BFD-5-SESSION_STATE_DOWN: BFD session 1090519112 to neighbor 10.22.101.1 on interface Vlan2201 has gone down. Reason: Control 
Detection Time Expired.
2021 Nov 17 07:37:30 TPLF0201 %BFD-5-SESSION_STATE_DOWN: BFD session 1090519107 to neighbor 10.22.101.2 on interface Vlan2201 has gone down. Reason: Control 
Detection Time Expired.
2021 Nov 18 05:09:12 TPLF0201 %ETHPORT-5-IF_DOWN_LINK_FAILURE: Interface Ethernet1/48 is down (Link failure)

Os alertas são eliminados em eno6/bd0 do labceed22 após a substituição do cabo.

[lab0200-smf/labceed22] cee# show alerts active summary
NAME UID SEVERITY STARTS AT SOURCE SUMMARY
---------------------------------------------------------------------------------------------------------------------------------------------------------
watchdog a62f59201ba8 minor 11-02T05:57:18 System This is an alert meant to ensure that the entire alerting pipeline is functional. This ale...