Skip to main content
NetApp Knowledge Base

What threshold value should be set for credit zero in Brocade switches?

Views:
57
Visibility:
Public
Votes:
0
Category:
fabric-interconnect-and-management-switches
Specialty:
brocade
Last Updated:

Applies to

  • Brocade Switches

Answer

  • Buffer credit zero counter increments are indications of credit latency. 
  • This indicates the number of times the port was unable to transmit frames because the transmit BB (buffer-to-buffer) credit was zero due to the delay in receiving R_RDY and not able to transmit the frames.
  • This counter incrementing is not necessarily a cause for concern but when you have a high number relative to the total number of frames transmitted (stat_ftx) then it can indicate degraded performance, such as congestion or a device affected by latency.
  • MAPS monitors the monitors the total latency over multiple window sizes with a different threshold for each time window. This allows MAPS to apply and monitor various levels of severity of latency at same time and at the VC level.
  • The monitoring window sizes and their corresponding thresholds are as follows:

70% of CRED_ZERO counter increment in 1 second.
50% of CRED_ZERO counter increment in 5 seconds.
30% of CRED_ZERO counter increment in 10 seconds.

  • The threshold values given above are default values and customizable using the FPI profiles.
  • However, it is  recommended to fix the issue with the device or devices causing this latency, and not change any buffer credit values instead.

Additional Information

  • When a violation occurs, the latency is reported as IO_PERF_IMPACT in the RASLog message; the message includes both the bandwidth loss amount and the corresponding time window. The message specifies the actual increment of the counter as a percentage.
    • The following example displays a typical RASLog entry with VC information for this condition. In this example, the bandwidth loss is 85% and the time window is 1 second.

2021/03/01-13:00:00, [MAPS-2070], 2692, SLOT 2 | FID 77, WARNING, Core_Switch, slot9 port19, E-Port 9/19, Condition=ALL_PORTS(DEV_LATENCY_IMPACT==IO_PERF_IMPACT), Current Value:[DEV_LATENCY_IMPACT, IO_PERF_IMPACT, (53.1% of 5 secs in VC: 9) ], RuleName=defALL_PORTS_IO_PERF_IMPACT, Dashboard Category=Fabric Performance
Impact.

 

  • The above MAPS message is not actually related to any buffer credit value, it's to do with with the tim_txcrd_z port stat counter, which indicates number of times that the port was unable to transmit frames because the transmit BB (buffer-to-buffer) credit was zero.
  • To understand 'tim_txcrd_z' events in porterrshow of Brocade switch please refer.
 
NetApp provides no representations or warranties regarding the accuracy or reliability or serviceability of any information or recommendations provided in this publication or with respect to any results that may be obtained by the use of the information or observance of any recommendations provided herein. The information in this document is distributed AS IS and the use of this information or the implementation of any recommendations or techniques herein is a customer's responsibility and depends on the customer's ability to evaluate and integrate them into the customer's operational environment. This document and the information contained herein may be used solely in connection with the NetApp products discussed in this document.