Technical Tip: SFP/DAC Cable Issues

[bqamar]

Description	This article describes how to troubleshoot issues with SFP/DAC module links. These modules come in many different varieties and types and therefore there are a number of things to consider when troubleshooting links using these Modules.
Scope	FortiSwitch supporting SFP modules .
Solution	Check Compatibility: It is very crucial to verify that the SFP module/DAC cable inserted into the FortiSwitch is supported by Fortinet.  The list of supported SFP modules can be found on thess links: FortiSwitch-Compatible Transceivers|FortiSwitch   Fortinet Transceivers DataSheet   Note: Not all Fortinet-labeled Modules are supported across all the FortiSwitches.   Verify if the module is detected and there are no Alarms:  Make sure that there is no 'Alarm' on the modules, and its getting sufficient Rx optical power, Voltage, and Current, as well as the operating temperature is within the range.  From FortiSwitch GUI/CLI:  Go under Monitor -> Modules. FortiSwitch CLI:   get switch modules summary    Portname   State    Type       Transceiver    RX  Vendor           Part Number      Serial Number    port49     INSERT  SFP/SFP+    10G-Base-LR    OK  FS               SFP-10GLR-31     F2040402729    port50     INSERT  SFP/SFP+    10G-Base-LR    OK  FS               SFP-10GLR-31     F2040402728  From FortiGate CLI: diagnose switch-controller  switch-info modules summary <S/N of FortiSwitch>  S224EPTF19-----9:   Portname   State    Type       Transceiver    RX  Vendor           Part Number      Serial Number    port49     INSERT  SFP/SFP+    10G-Base-LR    OK  FS               SFP-10GLR-31     F2040402729    port50     INSERT  SFP/SFP+    10G-Base-LR    OK  FS               SFP-10GLR-31     F2040402728  If State is not in Alarm and RX is 'OK', then skip the next step. Check Optical parameters on the module:  Run the Following command on the CLI to get more information about the operational parameters of the module: get switch modules status [port#]    Port(port50)  temperature 23.957031 C  voltage 3.293100 volts  laser_bias[0] 0.761600 mAmps  tx_power[0] -2.246809 dBm  rx_power[0] -2.926854 dBm  Check the operational limits of the module:  get switch modules limits [port#]  #sample output                       Alarm        ||       Warning              |   High   |   Low    ||   High    |  Low  temperature |  90.0000 | -45.0000 ||  85.0000 | -40.0000 C  voltage     |   3.8000 |   2.7000 ||   3.7000 |   2.8000 V  laser_bias  |  10.0000 |   0.0000 ||   9.0000 |   0.0100 mA  tx_power    |  -1.9997 | -10.0000 ||  -2.9999 |  -8.9997 dBm  rx_power    |   0.0000 | -23.9794 ||  -1.0002 | -23.0103 dBm  Compare the Parameters with the limits to find the reason why the module is in Alarm. Refer to this link for more details: Diagnostic monitoring interface module status  Check Speed/Duplex on both sides: Speed/Duplex setting must match on both sides. The default setting is 'auto-module' in which the modules negotiate the capabilities and sync on the maximum supported speed on each side. If the link does not come up on auto-module then both sides need to be configured on the same speed/duplex setting: If the FortiSwitch is not managed by the FortiGate(Standalone Switch):  config switch physical-por  edit [port#]  set speed [option] => Options listed below will vary depending on the module  auto        Auto-negotiation.  1000auto    Auto-negotiation (1G full-duplex only).   If the FortiSwitch is managed by the FortiGate and is currently online then follow the commands below, otherwise perform the configuration above in a), followed by the configuration below and using the same speed/duplex as used in a):  config switch-controller managed-switch      edit [S/N of FortiSwitch]          config ports              edit [port#]              set speed [option]              auto        Auto-negotiation.              1000auto    Auto-negotiation (1G full-duplex only).  Check the CRC errors on the ports: CRC or Cyclic Redundancy Check is a layer2 frame header property to validate if all the data arrived without corruption. The link may not come up if there are too many CRC errors being received: diagnose switch  physical-ports port-stats list  [port#]  Port23 Port Stats:             Rx Bytes:                                             0          Rx Packets:                                           0          Rx Discards:                                          0          Rx Errors:                                            0          Tx Errors:                                            0          Fragments:                                            0          Undersize:                                            0          Jabbers:                                              0          Collisions:                                           0          CRC Alignment Errors:                       0  The link should come up at this point: If other checks mentioned above have been validated, there might be some other issues present. Check error logs for any other reported issues.