What do CRC errors, runts, giants, input errors and output drops on an interface tell you?
Show the answer
Say this first
CRC errors are frames whose FCS did not match: a layer 1 problem such as bad cable, dirty fiber, a failing optic, EMI, or a duplex mismatch. Runts are frames under 64 bytes, giants frames over the maximum size (often an MTU or tagging mismatch). Input errors is the total of the input error types. Output drops are not errors on the wire: the egress queue was full, which means congestion. I clear the counters and watch which one grows.
Go deeper
A Catalyst interface with a physical problem
SW1# show interfaces GigabitEthernet1/0/24
GigabitEthernet1/0/24 is up, line protocol is up (connected)
Hardware is Gigabit Ethernet, address is 7c21.0e44.1a18 (bia 7c21.0e44.1a18)
MTU 1500 bytes, BW 1000000 Kbit/sec, DLY 10 usec,
Full-duplex, 1000Mb/s, media type is 10/100/1000BaseTX
Input queue: 0/2000/0/0 (size/max/drops/flushes); Total output drops: 0
5 minute input rate 412000 bits/sec, 61 packets/sec
5 minute output rate 389000 bits/sec, 58 packets/sec
1840122 packets input, 1306214450 bytes, 0 no buffer
0 runts, 0 giants, 0 throttles
4471 input errors, 4471 CRC, 0 frame, 0 overrun, 0 ignored
0 output errors, 0 collisions, 2 interface resetsCRC equals input errors, no runts, full duplex: a cabling or port problem, not a mismatch.
What each counter points to
| Counter | Likely cause |
|---|---|
| CRC | Cable, connector, optic, EMI, duplex mismatch (full side) |
| Runts | Collisions, duplex mismatch, faulty NIC |
| Giants | Jumbo frames into a 1500-byte port, MTU mismatch |
| Overrun, ignored, no buffer | The receiver could not keep up (hardware or buffer exhaustion) |
| Late collision | Duplex mismatch (half side), cable too long |
| Output drops | Congestion: speed mismatch, microbursts, oversubscription |
| Interface resets | Flapping link, keepalive failures |
Method
clear counters GigabitEthernet1/0/24, then watch for a few minutes. Old totals since boot tell you
nothing about now. On Catalyst, show interfaces counters errors gives a compact table per port.
They will ask next
CRC errors are on the switch port, but the server's NIC shows none. Which direction is bad?
CRC errors are counted by the receiver. The switch receiving bad frames means the path from the server to the switch is bad: the server's transmit side, the cable, or the switch's receive side. Swap the cable and patch first, then the port.
Traps
- Calling output drops a cabling problem. They are congestion.
- Reading lifetime counters without clearing them first.