RE: [nv-l] Cursed Cisco Trap Formats

2002-10-04 09:07:52
Subject: RE: [nv-l] Cursed Cisco Trap Formats
From: "Barr, Scott" <Scott_Barr AT csgsystems DOT com>
To: <nv-l AT lists.tivoli DOT com>
Date: Fri, 4 Oct 2002 08:07:52 -0500
Todd, those variables would not be blank, they would be the additional 3 
variables that James has been mentioning. Without knowing the total - which now 
appears possible, you can't tell what the data is in the trap. Like I said if 
the data is moving from field to field depending on the trap forrmat, I don't 
see how in code you could differentiate the type of data in the field. I 
considered your approach (in fact it was the only possible approach) but I 
could not get around not know what kind of data was in the variable - was it an 
ifType or ifIndex etc. When I know the total number of variables, I can pick 
the variables out based on which format they are. I know that if there are 5 
variables then the second one is ifType and the third one is ifReason etc. 

Thanks everyone - valuable input, I will be able to solve this. 

> -----Original Message-----
> From: Todd H. [mailto:netview AT toddh DOT net]
> Sent: Thursday, October 03, 2002 4:46 PM
> To: Barr, Scott
> Cc: nv-l AT lists.tivoli DOT com
> Subject: Re: [nv-l] Cursed Cisco Trap Formats
> "Barr, Scott" <Scott_Barr AT csgsystems DOT com> writes:
> > Very true - but I have no idea what the contents mean. if NVATTR_2
> > is "2" for ifType that is not the same if it is "2" for
> > ifLocalReason. Yes, blanks are detectable, but deciding which format
> > the trap is still won't be easy unless I can tell whether its a 3,
> > 4, or 5 type trap. 
> You know the formats for each case, and what sort of event you want to
> send or display you want to do, right? 
> Unless I've missed something, telling whether isn't testing for a 3 4
> or 5 trap is a simple matter of testing
> if NVATTR_1 2 and 3  are non blank. 
>         if 4&5 are blank or undefined, 
>                 do type 3 stuff using the defs you know for that one
>         fi
>         if (4nonblank and  5 blank/undefined)
>                 do type 4 stuff based on defs you know for that one
>         fi
>         if (4nonblank and 5 nonblank)
>                 do type 5 stuff
>         fi
> fi
> The syntax of that depends on your choice of scripting language. 
> So long as you can define any sort of footprint to distinguish the
> type of traps based on the varbinds available to you in the
> environment variables, you're all set. 
> -- 
> Todd H.
> http://www.toddh.net/