topic Re: CSS - Service transition trap in Application Networking

CSS - Service transition trap

jfoerster — Mon, 03 Jan 2005 09:42:02 GMT

Hi,

I've the feeling, That the CSS (webNS7.30) is not sending a service transition trap correctly. Normaly there are two values given (service name and service state).If I have a look at the MIB I see that these are the two defined objects. If I have a look at the received trap only one value stating the two variables is given (i.e.

local0log:Jun 13 02:03:31 atl-tme-linux snmptrapd[8334]: 10.90.14.233: Enterpris

e Specific Trap (1) Uptime: 1 day, 0:29:50.94, SNMPv2-SMI::enterprises.2467.1.15

.10.0 = STRING: "Service Transition - Service:sun State:down"

If I have a look at a server trap (i.e.

TRAP,10.10.10.10,1,6,3008

TRAPOID,public,.1.3.6.1.4.1.232

TIME, 1232523643,254778

OID,.1.3.6.1.2.1.1.5.0,[ABCD12345]

OID,.1.3.6.1.4.1.232.11.2.11.1.0,0

OID,.1.3.6.1.4.1.232.3.2.3.1.1.4.0.1,7

)

3 oids are given (the example is a trap of a compaaq server).

Is this a correct behaviour or in other words is this RFC compliant that two variables are given with only one OID?

TIA

Kind regards,

joerg

Re: CSS - Service transition trap

Gilles Dufour — Mon, 03 Jan 2005 14:13:02 GMT

this what is defined in the svcext.mib

apSvcTransitionTrap NOTIFICATION-TYPE

OBJECTS { apSvcTrapEventText }

STATUS current

DESCRIPTION

"A service transition trap is generated whenever the service

state transitions from ACTIVE -> DEAD, or DEAD -> ACTIVE."

::= { apSvcExtMibNotifs 1 }

apSvcTrapEventText OBJECT-TYPE

SYNTAX SnmpAdminString (SIZE(0..255))

So this trap will return a string.

The string can be whatever.

In this case it is the service name and status.

Up to you to extract the info that you need.

This is rfc compliant even if not very friendly for most of the existing tools.

Regards,

Gilles.

Re: CSS - Service transition trap

jfoerster — Mon, 03 Jan 2005 14:18:57 GMT

Hi Gilles,

thanks for the info.

I had the fear that this would be the answer. Is there a chance that this will be changed if necessary per feature request?

Kind Regards,

Joerg

Re: CSS - Service transition trap

Gilles Dufour — Tue, 04 Jan 2005 10:35:52 GMT

Joerg,

I know one customer who asked for this already.

Not sure if a feature request was submitted.

You can go through your sales representative to introduce the request as well.

Regards,

Gilles.

Re: CSS - Service transition trap

jfoerster — Wed, 05 Jan 2005 12:15:16 GMT

Hi Gilles,

having a look at your trapdefinition I found out, that this trap seems to be different one than the one I received as there is a difference in defintion given for this trap in svcEXT.mib and the definition you provided. In svcext.mib this trap is defined as trap-type and your defintions states notificatation type. Additionally the OID are differing...

Is this difference a result of my typo (I'm using WEBNS7.2 and not what I stated earlier 7.3) or is a different MIB I've not found so far.

TIA for clarification,

Regards,

Joerg

Re: CSS - Service transition trap

Gilles Dufour — Wed, 05 Jan 2005 15:13:39 GMT

Joerg,

the trap changed with 7.40 and this is the mib I actually checked because it was the one available to me quickly.

Checking 7.20, I see the following

apSvcTransitionTrap TRAP-TYPE

ENTERPRISE svcExt

VARIABLES { apSvcTrapEventText }

DESCRIPTION

"A service transition trap is generated whenever the

the service state transitions from ACTIVE -> DEAD, or

DEAD -> ACTIVE."

::= 1

Regards,

Gilles.

Re: CSS - Service transition trap

jfoerster — Thu, 06 Jan 2005 10:52:46 GMT

HI Gilles,

thanks for clarification. That's what I thought it to be.

Kind Regards,

Joerg