Callhome Link Events

[pcirne]

Hi, I just setup Callhome on a MDS 9148, it works fine with all the events except the port link events (up,down,failure,....)

I checked the Callhome profile and the syslogGroupPort is enabled, the link messages are logged in the syslog but I don't receive mail notifications....

Thank's for any suggestions on troubleshooting this issue!

Sent from Cisco Technical Support iPhone App

[dakester]

Hi Pedro,

It's probably these bugs.
CSCti92549    "Port logging level port link-failure 2" doesn't syslog as critical error
CSCti92353    IF_DOWN_LINK_FAILURE_CRIT not generated

If you are running 5.x, the fix will be in the next 5.x release.
   After upgrading to a release with the fix you will need to reconfigure the link-failure logging level to activate the fix.
   Config
   no logging level port link-failure crit
   logging level port link-failure crit
   copy run start

Regards,

David

[pcirne]

Hi Dakester,

The link failure events are logged:

2011 Apr  4 11:31:57 pordir03 %PORT-5-IF_DOWN_LINK_FAILURE: %$VSAN 1%$ Interface fc1/11 is down (Link failure)

2011 Apr  4 11:32:04 pordir03 %PORT-5-IF_DOWN_LINK_FAILURE: %$VSAN 1%$ Interface fc1/11 is down (Link failure)

2011 Apr  4 11:32:04 pordir03 %PORT-5-IF_UP: %$VSAN 1%$ Interface fc1/11 is up in mode F

But no CALLHOME is made for those events, other events like a power failure are notified by CALLHOME:

2011 Mar 31 16:48:41 pordir03 %PLATFORM-5-PS_STATUS: PowerSupply 2 current-status is PS_FAIL

2011 Mar 31 16:48:41 pordir03 %PLATFORM-2-PS_FAIL: Power supply 2 failed or shut down (Serial number QCS1430K1H5)

2011 Mar 31 16:48:41 pordir03 %CALLHOME-2-EVENT: POWER_SUPPLY_FAILURE

2011 Mar 31 16:48:41 pordir03 %CALLHOME-2-EVENT: FAN_FAILURE

The CALLHOME setup is made:

pordir03(config)# show callhome destination-profile profile full-txt-destination

full_txt destination profile information

maximum message size:500000

message-level:0

transport-method:email

email addresses configured:

DL-IT-SAN-Adm@nanium.com

url addresses configured:

alert groups configured:

system

environmental

linecard-hardware

supervisor-hardware

inventory

test

cisco-tac

syslog-group-port

RMON

I did what you have suggested:

pordir03(config)# no logging level port link-failure crit

pordir03(config)# logging level port link-failure crit

pordir03(config)# show logging level port

Facility        Default Severity        Current Session Severity

--------        ----------------        ------------------------

port                    5                       5

0(emergencies)          1(alerts)       2(critical)

3(errors)               4(warnings)     5(notifications)

6(information)          7(debugging)

pordir03(config)# show running-config all

!Command: show running-config all

!Time: Mon Apr  4 11:44:09 2011

version 5.0(4b)

license grace-period

logging level port link-failure 2

feature telnet

feature http-server

[dakester · Accepted solution]

Hi Pedro,

You hit the bug(s).

The Link Failure syslog messages are being logged as level 5.  This will not trigger a callhome.

2011 Apr  4 11:31:57 pordir03 %PORT-5-IF_DOWN_LINK_FAILURE: %$VSAN 1%$ Interface fc1/11 is down (Link failure)

pordir03(config)# show logging level port
Facility        Default Severity        Current Session Severity
--------        ----------------        ------------------------
port                    5                       5
2(critical)
5(notifications)

The Link Failure syslog messages need to be level 2 to trigger callhome.
The config has
version 5.0(4b)
logging level port link-failure 2

The config is not working. This is the bug.

9513-01# show port internal info all | i "syslog"
Global link failure syslog level = 2
    Link failure syslog level = 5  <---- config is not working at the interface level.
    Link failure syslog level = 5
    Link failure syslog level = 5

Options:
Wait for the next 5.x release.
Configure an snmp server to send emails for link failure traps.
Configure a syslog server to send emails for link failures.
Use RMON until you upgrade to the next 5.x release.
     Device Manager: Admin > Events > Theshold Manager.
       Select All interfaces, Link Fialures.
       Change the Sample(sec) from 10 to 30.
       Create (you may not get a confirmation, so don't click Create twice).
       To see the alarms, on bottom, click     More ... >  (top) 64 Bit Alarms

   For the callhome config, choose RMON for the full_txt callhome (default is all).
   The short_txt callhome may not contain the physical interface.

Regards,

David