amrith

updated

17 years ago

A

amrith

1 Rookie

•

76 Posts

0

1902

June 15th, 2009 02:00

How to find the response time from the Target to an initiator

How to find the response time from the target to the initiator in a Cisco MDS switch ?
Or in other words, How to find the time taken for data to travel from a target WWN port to the initiator WWN port.
Is fcping any useful in this case?
Thanks
  • healyj

    141 Posts

    1355

    0

    Posted November 28th, 2017 07:00

    Hi there,

    In our efforts to clean up the forum, we came across your question / statement.

    If the question / statement is still valid, not expired and you need an update please reach out again and we try to get it answered.

    As for now we set it to “answered.”

    Regards,

    Jim

  • 1355

    0

    Posted June 16th, 2009 12:00

    A fibre channel trace would allow you to calculate this.

    Thank you.
  • RRR

    6 Operator

    •

    5739 Posts

    1355

    0

    Posted June 17th, 2009 01:00

    I've never come accross a case where I questioned the respons times. Why would you want to know this ? I can only think of very long distances over long wave. Can you please tell us why you'd want to perform this test ?
  • amrith

    1 Rookie

    •

    76 Posts

    1355

    0

    Posted June 18th, 2009 08:00

    Appreciate your help.
    fctrace gives me the path in which the frame traverses from the switch it is ran to the destination WWN. How can I use this info to find out the time spent on each switch ?

    OSDC2E02# fctrace pwwn 50:xx:xx:80:15:27:9b:xx vsan 152
    Route present for : 50:06:0e:80:15:27:9b:74
    20:00:00:0d:ec:xx:49:00(0xfffc62)
    20:00:00:0d:ec:xx:2e:00(0xfffc84)
    20:00:00:0d:ec:xx:da:c0(0xfffcc2)
    20:00:00:0d:ec:xx:da:c0(0xfffcc2)
    20:00:00:0d:ec:xx:44:80(0xfffc86)

    RRR,

    A server having 2 HBAs gives slow performance on one path of the HBA and the other one work good.
    The storage response times are good indicating something is wrong in the path it takes to reach the storage port.

    I need to know where is the bottleneck causing the fabric which is causing delays.We have 16 switches and I am quite sure the problem is somewhere in the fabric.
  • RRR

    6 Operator

    •

    5739 Posts

    1355

    0

    Posted June 19th, 2009 05:00

    Just out of curiosity: what kind of values do we measure here ? ms, us, ns ?
  • sfuhrman

    25 Posts

    1355

    0

    Posted June 19th, 2009 14:00

    it would be in microseconds.I have typically measured 20-30us per switch on a Cisco director, and 2-3us per Brocade director (ingress to egress for an average frame). You need a fibre channel analyzer to be able to measure this, but I would agree with others in that latency inside the SAN is typically not an issue unless the design is bad and links are maxed out, etc.