Re: [c-nsp] prevent route entering fib

address prefix-list CUST_SEL_BLACKHOLE_32 match community 100 Erik - Oorspronkelijk bericht - Van: "Mark Tinka" <mark.ti...@seacom.mu> Aan: "Erik Klaassen" <e.klaas...@fr-ix.nl>, cisco-nsp@puck.nether.net Verzonden: Dinsdag 17 november 2015 10:44:39 Onde

Re: [c-nsp] prevent route entering fib

Hi, On Mon, Nov 16, 2015 at 07:35:48PM +0100, Pierre Emeriaud wrote: > 2015-11-16 19:26 GMT+01:00 Erik Klaassen <e.klaas...@fr-ix.nl>: > > > > What is the best way to prevent the exabgp route entering the fib? > > This may sound weird, but why not ann

[c-nsp] prevent route entering fib

I have a fastnetmon/exabgp instance to inject routes into the border router(7600 / 720-3b-xl) to trigger remote blackholing. Triggering a blackhole null route is easy. But now i want to implement selective blackholing (my upstream supports this) The key thing is not to null route in your own

Re: [c-nsp] Optical - rx power low warning

your loss over 20mile is a bit high. Assuming the other side is also sending at 2.4dbm, you receive -16.4 so your loss is 18.8. 20Mile is ~ 35 km. according to g652 standand loss @1550 should be 0.21dbm per km. So that would be 7,35. Add a cwdm mux/demux on both side is around 4dbm (2 per

Re: [c-nsp] Restrictions NetFlow v9 for IPv6

in this case the 7600 platform sub720-3bxl and ios 12.2sr* Erik - Oorspronkelijk bericht - Van: Jared Mauch ja...@puck.nether.net Aan: Erik Klaassen e.klaas...@fr-ix.nl Cc: cisco-nsp@puck.nether.net Verzonden: Woensdag 18 maart 2015 17:39:53 Onderwerp: Re: [c-nsp] Restrictions NetFlow

[c-nsp] Restrictions NetFlow v9 for IPv6

From the cisco netflow v9 guide: NDE v9 records for IPv6 do not contain Autonomous System (AS) numbers and prefix length information. Is this still the case? my src and dst AS ipv6 flow fields from my c7600 are 0. Is there some solution? Thanks, Erik

Re: [c-nsp] BGP dram confusion

Thanks all for the input, realy appriciate it. Of course a reboot freed the memory. Thanks, Erik - Oorspronkelijk bericht - Van: Gert Doering g...@greenie.muc.de Aan: Erik Klaassen e.klaas...@fr-ix.nl Cc: cisco-nsp@puck.nether.net Verzonden: Vrijdag 13 maart 2015 17:37:23 Onderwerp: Re

Re: [c-nsp] BGP dram confusion

I get the 3bxl has its limitations. I can live with the cpu and i dont need full bgp necessarily. But what i dont get is the current memory consumption. I filtered all the transits to only a default route. But still the memory stays at 90% full. I dont have soft-reconfig configured. Why is

[c-nsp] BGP dram confusion

Hi all, We use a 7600 with a sup720xl and we receive 3 full bgp tables, some partial transit and some peering. According to sh bgp sum bgp is using: ipv4 ~250MB ipv6 ~30MB But the 1GB dram is almost full. sh memory summary Head Total(b) Used(b) Free(b) Lowest(b)

Re: [c-nsp] BGP dram confusion

usage is the same! Thanks, Erik - Oorspronkelijk bericht - Van: Nick Hilliard n...@foobar.org Aan: Erik Klaassen e.klaas...@fr-ix.nl, cisco-nsp@puck.nether.net Verzonden: Woensdag 11 maart 2015 13:57:05 Onderwerp: Re: [c-nsp] BGP dram confusion On 11/03/2015 12:45, Erik Klaassen

Re: [c-nsp] Possible causes for fiber link flap?

we have not yet received full data from the splicing team. They're just saying it looks good. :-) 2.4dB loss over 3km is not in g652 spec. max dampening @1550 should be something like: (Length in Km * 0.25)+(Number of connectors * 0.5)+(number of splices * 0.1) Regards, Erik

Re: [c-nsp] Possible causes for fiber link flap?

2.4dB on a 3km link maybe well in your budget but its way to high for a 3km link. Must be something wrong with the splice. Swapping fibers seems a logic first step to narrow this down. Kind regards, Erik - Oorspronkelijk bericht - Van: Peter Rathlev pe...@rathlev.dk Aan: cisco-nsp

Re: [c-nsp] MACFLAP loop with l2pt

Solved by protecting the port with guard loop. Guard loop for the sudden loss of bpdu's on nondesignated ports. Weird this is not mentioned in the l2pt docs, very plausible to happen. Erik - Oorspronkelijk bericht - Van: Erik Klaassen e.klaas...@fr-ix.nl Aan: Aaron aar...@gvtc.com

Re: [c-nsp] MACFLAP loop with l2pt

No port-fast on any port. i am using r-pvst. Before is goes in fwd it is first learning indeed. should a port go fwd when its not receiving stp bpdu's? Is this a wrong implementation of l2pt for stp? Erik - Oorspronkelijk bericht - Van: Aaron aar...@gvtc.com Aan: Erik Klaassen e.klaas

[c-nsp] MACFLAP loop with l2pt

Hello, I have created a l2pt between sw1-gi0/1 and sw2-gi0/1 so both switches are logical connected. The te1/1 interface has the lowest cost so gi0/1 in sw2 goes in BLK state. When the connection between sw4 and the isp swich goes down, the link between sw2 and sw4 stays up but no bpdu's are