How to Enable Bottleneck detection feature in Brocade

Brocade has introduced a new Bottleneck detection feature in Brocade FOS v6.3.0 onwards.  The important thing to notice is that this feature doesn't require a license and supported on 4 Gbps, 8 Gbps, and 16 Gbps platforms. Bottleneck detection feature Monitors and reports latency and congestion bottlenecks on F_Ports and E_Ports. FOS v6.4.0 release introduced enhancements to improve the functionality of detecting device latency. 

Let's see how to enable it!

To check the status of the feature, whether it is enabled or disabled, you have to issue the command as below.

SANSW01:admin> bottleneckmon --status

Bottleneck detection - Disabled

To enable bottleneck monitoring  on  a  switch  with  alerts using  default  values for threshold and time, you have issue the command as below

SANSW01:admin> bottleneckmon --enable -alert

If the command is used without any detection parameters, default thresholds will be applied.The  congestion and Latency thresholds are expressed as a fraction between 0 and 1. The default value for congestion threshold is 0.8 and for latency, default value is 0.1

Setting a latency threshold of 0.1 and a time window of 300 seconds specifies that an alert should be  sent, when  10%  of  the  one-second samples over any period of 300 seconds were affected by latency bottleneck conditions.

Once bottleneck monitoring is enabled on a switch and -alert is  specified, if the ports on  the  configured  switch  experience latency  or congestion the  command  triggers  an SNMP alert

To fine tune the feature, --config operands is used along with following parameters,

-cthresh   Specifies  the severity threshold for congestion that triggers an alert.   The threshold indicates the percentage of one-second intervals affected  by  the  bottleneck  condition within the specified time window. 

-lthresh Specifies  the severity threshold for latency that triggers an alert. The  threshold  indicates  the percentage of one-second intervals affected by the bottleneck  condition  within the  specified  time window. 

-time Specifies  the  time  window  in seconds over which the bottleneck  conditions  is  computed and compared with the threshold. The default is 300 seconds.

-qtime Specifies  the  minimum  number  of   seconds  between  consecutive  alerts.  The default is 300 seconds. 

Eg:-

SANSW01:admin> bottleneckmon --config -alert -lthresh .2 -cthresh .9 -time 300 

To check the status
SANSW01:admin> bottleneckmon --status

Bottleneck detection - Enabled
==============================
Switch-wide sub-second latency bottleneck criterion:
====================================================
Time threshold                 - 0.800
Severity threshold             - 50.000
Switch-wide alerting parameters:
================================
Alerts                         - Yes
Latency threshold for alert    - 0.200
Congestion threshold for alert - 0.900A
veraging time for alert       - 300 seconds
Quiet time for alert           - 300 seconds

To display the number of ports affected by bottleneck conditions issue the command as below:

SANSW01:admin> bottleneckmon --show

To disable bottleneck monitoring on all ports in a chassis:

SANSW01:admin> bottleneckmon --disable