Large number of CRC errors occur on cluster ports

Applies to

• FAS2750
• FAS2720
• AFF C190A
• ONTAP 9
• Switchless Cluster

Issue

• CRC error counts multiple times on cluster port with ifconfig:

-- interface  e0b  (232 days, 2 hours, 21 minutes, 3 seconds) --

RECEIVE
 Total frames:      252m | Frames/second:      13  | Total bytes:     39644m
 Bytes/second:     1977  | Total errors:     1218  | Errors/minute:       0 
 Total discards:      0  | Discards/minute:     0  | Multi/broadcast:  1020k
 Non-primary u/c:     0  | CRC errors:       1212  | Runt frames:         0 
 Fragment:            0  | Long frames:         0  | Jabber:              3 
 Length errors:       3  | No buffer:           0  | Xon:                 0 
 Xoff:                0  | Pause:               0  | Jumbo:            1293k
 Noproto:             0  | Error symbol:      236  | Illegal symbol:    199 
 Bus overruns:        0  | Queue drops:         0  | LRO segments:      249m
 LRO bytes:       26713m
| LRO6 segments:       0  | LRO6 bytes:          0 
 Bad UDP cksum:       0  | Bad UDP6 cksum:      0  | Bad TCP cksum:       0 
 Bad TCP6 cksum:      0  | Mcast v6 solicit:    0 
TRANSMIT
 Total frames:      561m | Frames/second:      28  | Total bytes: 139g
 Bytes/second:     6980  | Total errors:        0  | Errors/minute:       0 
 Total discards:      0  | Queue overflow:      0  | Multi/broadcast:  1031k
 Xon:                 0  | Xoff:                0  | Pause:               0 
 Jumbo:            2565k | Cfg Up to Downs:     2  | TSO non-TCP drop:    0 
 Split hdr drop:      0  | Timeout:             0  | TSO segments:      402k
 TSO bytes:        4840m | TSO6 segments:       0  | TSO6 bytes:          0 
 HW UDP cksums:     154k | HW UDP6 cksums:      0  | HW TCP cksums:     556m
 HW TCP6 cksums:      0  | Mcast v6 solicit:    0 
DEVICE
 Mcast addresses:     3  | Rx MBuf Sz:       4096 
LINK INFO
 Speed:           10000M | Duplex:            full | Flowcontrol:       none
 Media state:     activ
e | Up to downs: 8

• EMS error message gives below output continuously every hour:

Tue Dec 01 04:04:26 JST [node 2: intr: netif.linkErrors:error]: Excessive link errors on network interface e0b. Might indicate a bad cable, switch port, or NIC, or that a cable connector is not fully inserted in a socket. On a 10/100 port, might indicate a duplex mismatch.
Tue Dec 01 05:04:27 JST [node 2: intr: netif.linkErrors:error]: Excessive link errors on network interface e0b. Might indicate a bad cable, switch port, or NIC, or that a cable connector is not fully inserted in a socket. On a 10/100 port, might indicate a duplex mismatch.
Tue Dec 01 06:04:27 JST [node 2: intr: netif.linkErrors:error]: Excessive link errors on network interface e0b. Might indicate a bad cable, switch port, or NIC, or that a cable connector is not fully inserted in a socket. On a 10/100 port, might indicate a duplex mismatch.
Tue Dec 01 07:04:27 JST [node 2: intr: netif.linkErrors:error]: Excessive link errors on network interface e0b. Might indicate a bad cable, switch port, or NIC, or that a cable connector is not fully inserted in a socket. On a 10/100 port, might indicate a duplex mismatch.
Tue Dec 01 08:04:28 JST [node 2: intr: netif.linkErrors:error]: Excessive link errors on network interface e0b. Might indicate a bad cable, switch port, or NIC, or that a cable connector is not fully inserted in a socket. On a 10/100 port, might indicate a duplex mismatch.
Tue Dec 01 09:04:29 JST [node 2: intr: netif.linkErrors:error]: Excessive link errors on network interface e0b. Might indicate a bad cable, switch port, or NIC, or that a cable connector is not fully inserted in a socket. On a 10/100 port, might indicate a duplex mismatch.

• Above error message does not stop even after replacing the cable between HA partner controller e0b port as an isolation.

::> event log show
Time                Node Severity      Event
------------------- ---------------- ------------- ---------------------------
1/5/2021 15:11:21 node 2    ERROR         netif.linkErrors: Excessive link errors on network interface e0b. Might indicate a bad cable, switch port, or NIC, or that a cable connector is not fully inserted in a socket. On a 10/100 port, might indicat
e a duplex mismatch.
1/5/2021 14:54:52   node 1    EMERGENCY     vifmgr.clus.linkdown: The cluster port e0b on node node 1 has gone down unexpectedly.
1/5/2021 14:54:52   node 2    EMERGENCY     vifmgr.clus.linkdown: The cluster port e0b on node node 2 has gone down unexpectedly.
1/5/2021 14:11:20   node 2    ERROR         netif.linkErrors: Excessive link errors on network interface e0b. Might indicate a bad cable, switch port, or NIC, or that a cable connector is not fully inserted in a socket. On a 10/100 port, might indicate a duplex mismatch.
1/5/2021 13:11:20   node 2    ERROR netif.linkErrors: Excessive link errors on network interface e0b. Might indicate a bad cable, switch port, or NIC, or that a cable connector is not fully inserted in a socket. On a 10/100 port, might indicate a duplex mismatch.
1/5/2021 12:11:19   node 2   ERROR         netif.linkErrors: Excessive link errors on network interface e0b. Might indicate a bad cable, switch port,
 or NIC, or that a cable connector is not fully inserted in a socket. On a 10/100 port, might indicate a duplex mismatch.
Press <space> to page down, <return> for next line, or 'q' to quit... q
6 entries were displayed.

• This is a switchless cluster configuration, target e0b port is a port that directly connects node 1 e0b to node 2 e0b as a cluster LIF ports.
• CRC errors count continues to increase after cable replacement and errors continue to report
• There is no difference in ports between each node connected to speed and duplex, and there is no use of transceivers such as SFP. 
• Cross connecting show errors remain with the port and do not follow the cable.