Bug in Call Home user defined destination profiles

Playing around with Call Home today and I seem to have hit a bug. I checked the Bug Database at Cisco and I see nothing that is logged there. Its possible this bug is marked “internal” like just about every other bug I have noted on this site. Basically if Cisco finds a bug first, they mark the bug id as internal and its not publicly viewable. If someone like a customer finds the bug first, then it becomes a public bug id. This is fine I guess, and probably determines some of the priority as to which bugs they will work on first. If users are finding the bugs, then its probably more serious than a bug that no one has ever called in about, but was only known internally. My only issue with this policy is that after a user calls in on bug that was previously marked internal only, they should then make it public, in my experience that doesn’t happen. Anyways, onto the bug……

Call Home allows you to create user defined destination profiles. You have to configure some basic stuff before being able to enable Call Home to begin with: SNMP ContactId, email-contact, phone-contact, and streetaddress. Once these are configured you can enable Call Home. Then to get notification via email you need to set an email from address and an smtp-server. After all this is done you can now configure a destination profile, which also means your configuring a destination email-address, and your all set.

You can use one of the built in destination profiles or a user defined one:

MDS1(config-callhome)# destination-profile ?
User defined destination profile name (Max Size – 31)
full-txt-destination Configure destination profile for plain txt message
short-txt-destination Configure destination profile for short txt message
XML-destination Configure destination profile for XML message

I have configured the built in ones, and those seem to work, at least using their default alert groups. However, when I configure my own user defined destination profile it does not seem to work. Here is the relevant information:

MDS1# show run | beg ^callhome
callhome
email-contact bfeeny@xxxxxx.com
phone-contact +1-561-214-2470
streetaddress 3250 W Commercial Blvd, Ste 360, Oakland Park, Fl 33309
destination-profile TEST
destination-profile TEST format full-txt
destination-profile TEST message-size 0
destination-profile TEST email-addr bfeeny@xxxxx.com
destination-profile TEST alert-group RMON
transport email smtp-server exchange.ctiusa.com port 25
transport email from bfeeny@yyyyy.com
enable
alert-group RMON user-def-cmd “show run | inc rmon”

I masked out my email addresses so that web crawlers don’t add me to a zillion email lists. But I can tell you that the SMTP configuration is correct, as it works fine with the built in destination profiles. I of course have a valid ip name-server set. I have RMON events triggering almost continuously, yet no email is ever sent:

MDS1# show callhome destination-profile profile TEST
TEST destination profile information
maximum message size:0
message format:full-txt
message-level:0
email addresses configured:
bfeeny@xxxxxxx.com

alert groups configured:
RMON

I will update this blog post with the outcome of the TAC case I have opened, but most likely this is a bug. I am using 3.3(5) and have not tested this in any other versions.

This entry was posted in Call Home, CCIE Storage and tagged , , , , . Bookmark the permalink.

4 Responses to Bug in Call Home user defined destination profiles

  1. Great, I never knew this, thanks.

  2. Caio Oliveira says:

    Feeny, is there a way to show in the CLI the levels from the message-level on CallHome?
    I’m asking that because if a question especify like critical system alert-group, I have to configure it in the Device Manager and that I discover that 6 is the message-level for critical
    Regards,
    Caio

    • brian says:

      Not sure if you can re-phrase, I am not sure I understand what you are saying. Call home severity levels are of course different than syslog severity levels. Are you saying you want to verify the severity level of an alert group?

  3. Caio Oliveira says:

    Sorry, I will re-phrase.
    I would like to know if there is a command in the CLI to discover the destination-profile message-level.
    Like this destination-profile X message-level 0 is debug, 1 is normal, 2 is notif… and so on…

    On the CLI, I canĀ“t see that 0 is debug, 1 is normal, etc but int the DeviceManager I can set Normal, Debug,Critical, etc…

Leave a Reply