An administrator receives an alert in Prism indicating that interface eth2, on an AHV host is receiving many CRC errors. After togging into the problematic host, the following command is run to show the indicated output:
What is causing this issue?
Ref:
rx_crc_errorsare caused either by faults in layer 1 (in the past, we have seen failed twinax cables and incorrect types of fibre being used), or issues with jumbo frames on the network. In an environment with 10 Gig switches that use cut-through forwarding (Cisco Nexus, Arista, Cisco devices using IOS default to Store and Forward switching), any packets that come into the switch will get forwarded out the destination interface once the switch has read the destination MAC address. If that packet has an MTU over what is configured on the interface, it will cut off the packet at the designated MTU, causing the server to receive a malformed packet, which will throw a CRC error.
If you have a layer 1 issue, you will seerx_crc_errors,not on all but one or two nodes.
Currently there are no comments in this discussion, be the first to comment!