Thank you all for the suggestions you gave.
I understand that "yellow is bad", and that RFI could be very usefull
in some network topology. However in my case, it make sense that an
interface down is not as severe as a router down: some sites are
connected with 2 serial lines to 2 different backbone routers for
redundancy. Same reasons plays against James suggestion of changing
the default propagation rule. So I think I am still going to turn
off RFI here.
Thierry
On 28.06.2001 20:39:15 "Scott Barr" wrote:
>Two additional comments
>
>RFI is most useful for routers sitting in front of server farms. We page
>when a server is unreachable. If a router goes down, and the entire server
>farm is down, you would then page on each and every node. RFI prevents the
>node down traps from the devices sittting behind the router. It also
>suppresses the pings until the router is back up.
>
>The second point I would make is that the scenario you mentioned assumes a
>serial link - i.e. a router at the end of a phone line. In our
environment,
>we have routers connected between token ring and ethernet segments and in
>that case, you will not get a local router sending a link down trap. In
the
>serial link example, the local router would send a link down trap.
>
>-----Original Message-----
>From: owner-nv-l@tkg.com [mailto:owner-nv-l@tkg.com]On Behalf Of Leslie
>Clark
>Sent: Thursday, June 28, 2001 6:41 AM
>To: IBM NetView Discussion
>Subject: Re: [NV-L] RFI not very useful?
>
>
>By way of explanation, I think it was assumed that when Router B went
>down, the adjoining interface on Router A would stay up, making it
>more clear just where the problem was. Lately I have been seeing that
>in most networks, Router A shows its interface as down as well. So
>the event you need to respond to is Router Marginal. That will not be the
>case in all kinds of networks. Les is right: Any interface outage on a
>router must be handled as serious. That means of course that you must
>clean up all of those miscellaneous down interfaces that are making
>routers yellow on a regular bases: deconfigure old unused interfaces,
>convert to snmp polling for those which you can poll but not ping,
>unmanage BRIs, etc. So your map is all green under normal circumstances,
>and any yellow is to be investigated.
>
>Cordially,
>
>Leslie A. Clark
>IBM Global Services - Systems Mgmt & Networking
>Detroit
>
>
>James Shanks/Raleigh/IBM@IBMUS@tkg.com on 06/26/2001 02:51:56 PM
>
>Please respond to IBM NetView Discussion <nv-l@tkg.com>
>
>Sent by: owner-nv-l@tkg.com
>
>
>To: IBM NetView Discussion <nv-l@tkg.com>
>cc:
>Subject: Re: [NV-L] RFI not very useful?
>
>
>
>Another alternative mightb to change the propagation rules for your map
>(use Edit --> Modify/Describe --> Map)
>and chnage from the default to either "Propagate Most Critical" so that
any
>red causes the parent to go red, or "Propagate At Threshold Value"
>and see if you like the results better.
>
>James Shanks
>Team Leader, Level 3 Support
>Tivoli NetView for UNIX and NT
>
>
>"Les Dickert" <lesdickert@HOTMAIL.COM>@tkg.com on 06/26/2001 02:25:17 PM
>
>Please respond to IBM NetView Discussion <nv-l@tkg.com>
>
>Sent by: owner-nv-l@tkg.com
>
>
>To: nv-l@tkg.com
>cc:
>Subject: Re: [NV-L] RFI not very useful?
>
>
>
>Probably nothing to do to "improve" the
>situation other than how you deal with
>them.
>
>We treat yellow routers as bad and go
>drilling for the red. Usually it is
>an interface down, which is bad.
>
>Also, you should be getting an "Interface
>Down" event, which is usually also bad,
>especially if they aren't supposed to go
>down.
>
>Main message: Yellow is bad.
>
>Les Dickert
>Verisign Consulting
>
>
>
>>From: thierry.van-mol@CIEV.vd.ch
>>Reply-To: IBM NetView Discussion <nv-l@tkg.com>
>>To: " - *nv-l@tkg.com" <nv-l@tkg.com>
>>Subject: [NV-L] RFI not very useful?
>>Date: Tue, 26 Jun 2001 18:42:05 +0200
>>
>>
>>I found out a behaviour of RFI which let me think that I might need
>>to turn it off on our network. Situation is like this:
>>
>>
>>
>>
>>Netview --------- Router A -------------- Router B ------- Subnet B
>> Serial link
>> PPP
>>
>>
>>
>>Serial link is configured with a IP subnet of 30 bits (255.255.255.252
>>mask).
>>
>>Now, when Router B is switch off completely (power failure):
>>
>>- Subnet B and Router B becomes white (unreachable)
>>
>>- Router A becomes yellow
>>
>>- several interface, segment and network down events. A router marginal
>> event for both router A and router B
>>
>>- no Node down or Router down event for Router B is sent. I don't even
get
>>a
>> Router unreachable event (?)
>>
>>This behavior match the description of RFI in the doc. Indeed all subnets
>>to which Router B is connected are unreachable (since the serial
interface
>>on router A went down, the subnet of the serial link has no more
reachable
>>host). But I think that for the operator, the representation is
>misleading:
>>this is a very serious outage in the network, and the root fault is
Router
>>B.
>>But I don't see anything becoming red on the Home submap, and no major
>>event like "Node down" comes in. Only a marginal event for router A.
>>
>>Does anybody has a suggestion for improving this situation?
>>
>>
>>Thierry
>>
>>
>>_________________________________________________________________________
>>NV-L List information and Archives: http://www.tkg.com/nv-l
>
>_________________________________________________________________
>Get your FREE download of MSN Explorer at http://explorer.msn.com
>
>_________________________________________________________________________
>NV-L List information and Archives: http://www.tkg.com/nv-l
>
>
>_________________________________________________________________________
>NV-L List information and Archives: http://www.tkg.com/nv-l
>
>
>_________________________________________________________________________
>NV-L List information and Archives: http://www.tkg.com/nv-l
>
>_________________________________________________________________________
|