SAN Latency

Hello,

Connectrix Manager and ECC have the options to collect performance graphs and reports. Hope that answers your query.

Thanks,

Ali

Hello,

As mentioned above Connectrix Manager and ECC are excellent solutions.

However there is a way to do it in CLI as well through bottleneck detection.

A bottleneck is a port in the fabric where frames cannot get through as fast as they should. In other words, a bottleneck is a port where the offered load is greater than the achieved egress throughput. Bottleneck detection does not require a license.

Bottlenecks are reported through RASlog alerts. You can set alert thresholds for the severity and duration of the bottleneck. You can also use a CLI command to display a history of bottleneck conditions on a port. A history is maintained for a maximum of three hours for each port.

Kindly find below examples on how to enable and display bottleneck detection.

Enabling bottleneck detection on a port:

• When you enable bottleneck detection, you also determine whether RASlog alerts are to be sent when the bottleneck conditions at a port exceed a specified threshold.
  

• Displaying bottleneck history:

  • You can use the following procedure to display a history of bottleneck conditions, for up to three hours, on a port.
    
• Connect to the switch to which the target port belongs and log in as admin.
  
• Enter the bottleneckmon --show command to display a history of the bottleneck severity for a specific port.
  
• Example of displaying the bottleneck history for port 3 in 5-second windows over a period of 30 seconds
  

Ayman,

is there an equivalent process for Cisco switches (other then looking at Cisco performance manager)

Thanks

Hello,

I am unaware of a similar setup for cisco (other than the ones provided by Device manager and Performance manager).

However, The basic steps for troubleshooting port issues on a cisco switch using CLI would be:

• Use the show interface command to determine if the host HBA and the storage port can provide link level connectivity to their respective switches.
  NPI1# show interface fc2/5 status
  fc2/5 is down (Offline)
  NPI2# show interface fc2/5 status
  fc1/5 is up Port mode is F
  

• If the port is down and offline, use the no shutdown command to bring the port online.
  NPI1# config t
  NPI1(config)# interface fc 2/5
  NPI1(config-if)# no shutdown
  

• Repeat first bullet to determine if the port is online.
  If either of the ports fails to remain in the online state, then you may have a faulty GBIC, cabling or HBA/subsystem port.
  

• If both ports are online, use the show flogi command to verify that the Fibre Channel ports for the host and storage have performed a fabric login (FLOGI) and are communicating with their respective switches.
  NPI1# sh flogi
  INTERFACE VSAN FCID PORT NAME NODE NAME
  -------------------- ---- -------- ----------------------- -----------------------
  fc2/5 1 0x7e0200 21:00:00:e0:8b:08:d3:20 20:00:00:e0:8b:08:d3:20
  fc2/7 1 0x7e0300 20:00:00:e0:69:41:98:93 10:00:00:e0:69:41:98:93
  .....
  

• If you do not see the ports in the show flogi output, use the debug flogi even interface command to isolate the FLOGI issue.
  NPI1# debug flogi event interface fc2/5
  

• If the ports are in the show flogi output, use the show fcns database command to verify that the assigned FC ID during FLOGI exists in the name server database.
  NPI2# show fcns database
  --------------------------------------------------------------------------
  FCID TYPE PWWN (VENDOR) FC4-TYPE:FEATURE
  --------------------------------------------------------------------------
  0x9f0100 N 50:06:04:82:c3:a0:98:5c (EMC) scsi-fcp:target 250
  0x7e0200 N 21:00:00:e0:8b:08:d3:20 (QLogic) scsi-fcp:init
  
• At this point the HBA and subsystem ports have successfully established link level connectivity and each one can communicate with its locally attached switch in the fabric. The next step is to verify zone membership.
  

Also you could configure call home features to report on degraded performance but the setup is a bit too complex to explain here in the forum.

I hope this answers your question.

Have a great day.

Regards,

Ayman

The only latency related command I'm aware of on the Cisco switches is

fcping. Of course that's adhoc and not active monitoring.

Thanks Ayman  ..i was hoping for something lower level, for example new NX-OS allows to monitor port group oversubscribtion ..so something of that nature.

Thanks

Hello dynamox.

It's a definite that the GUI provides latency graphs to monitor performance, however I am unaware of a similar functionality in CLI.

I apologize for the inconvenience.

Regards,

Ayman

no need to apologize, thanks for looking into it

If not, could you please mark the question as Answered (helpful or correct as applicable)?

sorry but i can't, this question was asked by somebody else.

Glad I could help.

If you have any other questions/problems please let me know.

If not, could you please mark the question as Answered (helpful or correct as applicable)?

Have a great day.

Regards,

Ayman