From: sabrina pittarel (sabri_esame@yahoo.com)
Date: Sat Aug 26 2006 - 22:18:18 ART
Yep,
you are correct. I see the same behavior.
However that's not true for all traps. PIM traps are
sent to only the NMS configured to receive them.
This behavior is quite interesting...does anybody have
any more insight on this?
Sabrina
--- allboutcisco <frenzeus@streamyx.com> wrote:
> Hi Ivan,
>
> I believe by removing the global traps, no traps
> will be generated at all,
> even if snmp-server hosts were configured with
> specific traps to send to.
> Tried that in my lab.
>
> Thanks.
>
> -K
>
> ----- Original Message -----
> From: "Ivan" <ivan@iip.net>
> To: <ccielab@groupstudy.com>; "allboutcisco"
> <frenzeus@streamyx.com>
> Sent: Wednesday, August 23, 2006 12:31 AM
> Subject: Re: Sending specific SNMP traps to NMS host
>
>
> > Once you have this configured. HSRP and OSPF traps
> will send to all
> > snmp-hosts. Try to remove this.
> >> snmp-server enable traps hsrp
> >> snmp-server enable traps ospf
> >
> >
> > On Tuesday 22 August 2006 19:50, allboutcisco
> wrote:
> >> Hi group,
> >>
> >> I've been doing some verification on my config
> for snmp. I configured 2
> >> snmp host on the router to forward traps to,
> specifically 1 which
> >> receives
> >> only hsrp and the other no specific traps
> specified. Then i enable traps
> >> globally for hsrp & ospf. I have always thought
> or at least assumed that
> >> by
> >> specifying "snmp-server host 141.1.77.100 version
> 2c CISCO hsrp" that
> >> the
> >> router will only send hsrp-related traps to the
> NMS 141.1.77.100; but it
> >> proves otherwise when i did a debug snmp packet.
> >>
> >> My config as follows:
> >> snmp-server host 141.1.7.100 CISCO tty
> >> snmp-server host 141.1.77.100 version 2c CISCO
> hsrp
> >> snmp-server enable traps hsrp
> >> snmp-server enable traps ospf
> >>
> >> My findings are:
> >> When an event takes place for hsrp & ospf, snmp
> traps are generated,
> >> queued
> >> & finally sent (well at least from what the
> debugs say) to both
> >> 141.1.7.100
> >> & 141.1.77.100. To me, it doesn't seem that by
> tying in a specific trap
> >> to
> >> the snmp host ONLY that specific trap will be
> sent to it.
> >>
> >> My debug output (debug snmp packets & debug ip
> packet detail)
> >>
> >> Rack1R3#
> >> *Mar 1 15:02:59.707: SNMP: Queuing packet to
> 141.1.7.100
> >> *Mar 1 15:02:59.707: SNMP: V1 Trap, ent
> cHsrpMIBNotificationPrefix, addr
> >> 141.1.37.3, gentrap 6, spectrap 1
> cHsrpGrpEntry.15.3.1 =
> >> NO_SUCH_INSTANCE_EXCEPTION
> >> *Mar 1 15:02:59.779: SNMP: Queuing packet to
> 141.1.77.100
> >> *Mar 1 15:02:59.783: SNMP: V2 Trap, reqid 22,
> errstat 0, erridx 0
> >> sysUpTime.0 = 925530
> >> snmpTrapOID.0 = cHsrpMIBNotifications.1
> >> cHsrpGrpEntry.15.3.1 =
> NO_SUCH_INSTANCE_EXCEPTION
> >> *Mar 1 15:02:59.959: SNMP: Packet sent via UDP
> to 141.1.7.100
> >> *Mar 1 15:02:59.959: IP: tableid=0, s=141.1.37.3
> (local), d=141.1.7.100
> >> (FastEthernet0/0), routed via FIB *Mar 1
> 15:02:59.959: IP: s=141.1.37.3
> >> (local), d=141.1.7.100 (FastEthernet0/0), len 93,
> sending *Mar 1
> >> 15:02:59.959: UDP src=52433, dst=162
> >> Rack1R3#
> >> *Mar 1 15:03:00.035: SNMP: Packet sent via UDP
> to 141.1.77.100
> >> *Mar 1 15:03:00.035: IP: tableid=0, s=141.1.37.3
> (local), d=141.1.77.100
> >> (FastEthernet0/0), routed via FIB *Mar 1
> 15:03:00.035: IP: s=141.1.37.3
> >> (local), d=141.1.77.100 (FastEthernet0/0), len
> 118, sending *Mar 1
> >> 15:03:00.035: UDP src=58747, dst=162
> >> Rack1R3#
> >>
> >>
> >> Appreciate any inputs on this. Thanks.
> >>
> >> -K
> >>
> >>
>
This archive was generated by hypermail 2.1.4 : Fri Sep 01 2006 - 15:41:58 ART