如何对CRC以太网物理链路问题进行故障排除？
适用场景
- CRC错误故障排除
- 物理链路
问题描述
在接收端口上报告CRC错误、因此发生原因通常在到达之前就已找到。
- EMS：
[Node-01: intr: netif.linkErrors:error]: Excessive link errors on network interface e0a. 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.
[Node-01: vifmgr: callhome.clus.net.degraded:alert]: Call home for CLUSTER NETWORK DEGRADED: CRC Errors Detected - High CRC errors detected on port e0a node Node-01
[Node-02: vifmgr: vifmgr.cluscheck.crcerrors:alert]: Port e4a on node Node-02 is reporting a high number of observed CRC errors.
- 运行状况警报：
::> system health alert show Node: node_name Resource: Ethernet1/29/1 Severity: Major Indication Time: Mon Jan 04 16:15:33 2021 Suppress: false Acknowledge: false Probable Cause: The percentage of inbound packet errors of switch interface "switch_name(XXXXXXXXXXX)/Ethernet1/29/1" is above the warning threshold. Possible Effect: Communication between nodes in the cluster might be degraded. Corrective Actions: 1) Migrate any cluster LIF that uses this connection to another port connected to a cluster switch. For example, if cluster LIF "clus1" is on port e0a and the other LIF is on e0b, run the following command to move "clus1" to e0b: "network interface migrate -vserver vs1 -lif clus1 -sourcenode node1 -destnode node1 -dest-port e0b" 2) Replace the network cable with a known-good cable. If errors are corrected, stop. No further action is required. Otherwise, continue to Step 3. 3) Move the network cable to another port on the node (if available). Migrate the cluster LIF to the new port. If errors are corrected, contact technical support to troubleshoot the original node port. Otherwise, continue to Step 4. 4) Move the network cable to another available cluster switch port. Migrate the cluster LIF back to the original port. If errors are corrected, contact technical support to troubleshoot the original switch port. If errors persist, contact technical support for further assistance.
ifstat：
::> system node run -node <node> -command ifstat <port> -- interface e0a (62 days, 23 hours, 58 minutes, 45 seconds) -- RECEIVE Total frames: 81107m | Frames/second: 14901 | Total bytes: 387t Bytes/second: 71201k | Total errors: 1396k | Errors/minute: 15 Total discards: 0 | Discards/minute: 0 | Multi/broadcast: 1646k Non-primary u/c: 0 | CRC errors: 1396k | Runt frames: 0 Fragment: 0 | Long frames: 10 | Jabber: 0 Length errors: 108 | No buffer: 0 | Xon: 0 Xoff: 0 | Pause: 0 | Jumbo: 45712m Noproto: 0 | Error symbol: 0 | Illegal symbol: 0
- 单个集群端口报告
ifconfig -vvv输出mW低：
::> system node run -node <nodename> -command ifconfig -vvv … e0b: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> metric 0 mtu 9000 uuid: 0320a80b-caa3-11eb-b14a-d039ea306760 ... RX: 0.06 mW (-12.13 dBm) TX: 0.55 mW (-2.59 dBm)