On 06/01/2010 11:31 AM, Javier wrote:
> Risto,
>
> yes, you´re right. I receive 3 different traps but from 3 different
> generic OID's but that´s not important. For example
>
> <generic OID1> .1.2.3 A C
> <generic OID2> .1.2.3 A D
> <generic OID3> .1.2.3 B D
>
> and each one of them, brings the raw data:
>
> <generic OID1> .1.2.3 ="OK"
> <generic OID1> .1.2.3.5="234234"
> <generic OID1> .1.2.3.6.7="UP"
>
> As you can see, some of them have more numbers in the OID, so how can i
> process all of them independently of the amount of numbers in the OID?.
> If is it not easily acomplished anyway, i guess i can receive the
> relevant data from the same amount of numbers in the OID.

it should be fairly easy to write a regular expression that handles the 
variable amount of numbers. For example, [\d.]+=(\S+) matches any 
sequence of numbers and dots that is followed by '=', and it also 
assigns the value that follows '=' to a variable $1.

>
> What about to use Context to save every value from every parameter of
> the trap and when its a data change, write it in another log? (maybe
> that perl script in http://snmptt.sourceforge.net/docs/snmptt.shtml#SEC
> to trap back). Is it possible??.

Indeed, you could use contexts for memorizing event correlation state, 
but I'd recommend to take advantage of the 'desc' field of a rule.

BR,
risto

>
>
> Thank you very much
>
>
> 2010/6/1 Risto Vaarandi <[email protected]
> <mailto:[email protected]>>
>
>     Javier,
>
>     if each trap has several parameters that can have different values, and
>     you want to correlate events with the *same* set of values, this can be
>     easily accomplished by setting the 'desc' parameter accordingly.
>
>     Suppose that the trap with OID .1.2.3 can have two parameters:
>
>     <generic OID> .1.2.3 A C
>     <generic OID> .1.2.3 A D
>     <generic OID> .1.2.3 B D
>
>     With this example, it is obvious that the following rule will react only
>     to the first trap:
>
>     type=SingleWithSuppress
>     ptype=RegExp
>     pattern=<generic OID> (\S+)
>     desc=estado $1
>     action=shellcmd /home/javier/send.sh
>     window=300
>
>     However, if you rewrite the rule as follows, each trap .1.2.3 with a
>     *different* set of parameter values will be correlated by a different
>     event correlation operation (in other words, if a trap with a new set of
>     parameter values comes in, it is written to the log and then suppressed
>     for 5 minutes):
>
>     type=SingleWithSuppress
>     ptype=RegExp
>     pattern=<generic OID> (\S+) (\S+) (\S+)
>     desc=estado $1 $2 $3
>     action=shellcmd /home/javier/send.sh
>     window=300
>
>     The key to the solution is to define the 'desc' parameter correctly,
>     since 'desc' defines the scope of event correlation operations. Similar
>     question has been asked many times in this list before, and since this
>     is one of the fundamentals of SEC, please have a look at the relevant
>     section of the man page:
>     http://simple-evcorr.sourceforge.net/sec.pl.html#lbAV
>     This section explains the relation between rules and event correlation
>     operations, and how the 'desc' parameter influences this.
>
>     BR,
>     risto
>
>     On 06/01/2010 10:32 AM, Javier wrote:
>      > Hi,
>      >
>      > well, that trap comes from a device with several parameters. I
>     receive 3
>      > different traps, some common parameters to all traps and others
>     only for
>      > each trap. A change in the trap is when I receive a different data in
>      > any parameter from the last same trap.
>      >
>      > I´ve been searching documentation and maybe i could use Context
>     to save
>      > the data from each parameter and then show it later if it
>     changes, but i
>      > don´t know exactly how i can do that. Is it possible?. Any
>     suggestion?
>      >
>      > thanks in advance
>      >
>      >
>      > 2010/5/31 Risto Vaarandi <[email protected]
>     <mailto:[email protected]> <mailto:[email protected]
>     <mailto:[email protected]>>>
>      >
>      >     How do you define a change in the trap?
>      >     br,
>      >     risto
>      >
>      >
>      >     --- On Mon, 5/31/10, Javier <[email protected]
>     <mailto:[email protected]>
>      > <mailto:[email protected] <mailto:[email protected]>>> wrote:
>      >
>      >     From: Javier <[email protected] <mailto:[email protected]>
>     <mailto:[email protected] <mailto:[email protected]>>>
>      >     Subject: [Simple-evcorr-users] Can SEC help me ??
>      >     To: "simple-evcorr-users"
>     <[email protected]
>     <mailto:[email protected]>
>      > <mailto:[email protected]
>     <mailto:[email protected]>>>
>      >     Date: Monday, May 31, 2010, 8:15 PM
>      >
>      >     Hi,
>      >
>      >     i need to make a persistance correlation and i´m not sure if
>     SEC can
>      >     help me. It should goes like this:
>      >
>      >     I receive traps with a determinate OID in a log file as raw data,
>      >     then i write the output to a DB to show it later as an alarm.
>      >
>      >
>      >     I want to show the first coincidence and i want to correlate the
>      >     event to show only when its a change in the trap during a time
>      >     threshold and if there´s a change in that same event, reset that
>      >     time threshold and of course show the changes. I prefer do it in
>      >     memory and not in DB...
>      >
>      >
>      >     Well, is it possible to do that with SEC? and if yes...
>     HOW??? maybe
>      >     using several rules in the same conf file??
>      >
>      >
>      >     thanks in advance
>      >
>      >
>      >
>      >
>      >     -----Inline Attachment Follows-----
>      >
>      >
>     
> ------------------------------------------------------------------------------
>      >
>      >
>      >     -----Inline Attachment Follows-----
>      >
>      >     _______________________________________________
>      >     Simple-evcorr-users mailing list
>      > [email protected]
>     <mailto:[email protected]>
>      > <mailto:[email protected]
>     <mailto:[email protected]>>
>      > https://lists.sourceforge.net/lists/listinfo/simple-evcorr-users
>      >
>      >
>      >
>      >
>      >
>      >
>      >
>      >
>     
> ------------------------------------------------------------------------------
>      >
>      >
>      >
>      >
>      > _______________________________________________
>      > Simple-evcorr-users mailing list
>      > [email protected]
>     <mailto:[email protected]>
>      > https://lists.sourceforge.net/lists/listinfo/simple-evcorr-users
>
>
>     
> ------------------------------------------------------------------------------
>
>     _______________________________________________
>     Simple-evcorr-users mailing list
>     [email protected]
>     <mailto:[email protected]>
>     https://lists.sourceforge.net/lists/listinfo/simple-evcorr-users
>
>
>
>
> ------------------------------------------------------------------------------
>
>
>
>
> _______________________________________________
> Simple-evcorr-users mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/simple-evcorr-users


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

_______________________________________________
Simple-evcorr-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/simple-evcorr-users

Reply via email to