Brocade B6505 Link CRC Error and Optical Attenuation Troubleshooting Case

Troubleshooting Guide for Optical Signal Degradation on Brocade B6505 Switches

Fault Symptoms

Two Lenovo SR650 servers are connected to a Lenovo DE4000H storage array via a Lenovo B6505 Fibre Channel switch. The two servers are configured in a High Availability (HA) cluster. During HA failover, database access on the servers was observed to be slow, and sometimes inaccessible.

Solution

1. Reviewed the server LXCC logs; no hardware errors were found. Additionally, upgraded the server hardware drivers and microcode to the latest versions, but the issue persisted.

2. Analyzed the Lenovo DE4000H storage logs; no errors were detected.

3. When inspecting the Lenovo B6505 Fibre Channel switch using the command `porterrshow`, it was observed that `Crc_err` values were present and continuously increasing. After troubleshooting the fibre cables, optical modules, and HBA cards, the root cause was identified as a faulty optical module on the Fibre Channel switch. Following the replacement of the optical module, monitoring showed that the `Crc_err` value remained at 0 with no increase, resolving the issue.

Notes

`Crc_err` represents the count of frames with CRC checksum errors. If this error count continues to rise, there is likely a physical fault on the link. Check the fibre cables, optical modules on the Fibre Channel switch, HBA cards, etc.