[j-nsp] Re: juniper-nsp Digest, Vol 12, Issue 2
jgrewal
j.grewal at in.spectranet.com
Tue Nov 4 22:34:44 EST 2003
Can you try configuring interface as next-hop rather then giving ip address
. May be your both ip addresses reachable from both the interfaces so using
only one link.Also pls check in Juniper load balancing is per flow not per
packet so your outgoing traffic load balancing in many cases is influenced
by your incoming traffic.
J.Grewal
----- Original Message -----
From: <juniper-nsp-request at puck.nether.net>
To: <juniper-nsp at puck.nether.net>
Sent: Sunday, November 02, 2003 10:32 PM
Subject: juniper-nsp Digest, Vol 12, Issue 2
> Send juniper-nsp mailing list submissions to
> juniper-nsp at puck.nether.net
>
> To subscribe or unsubscribe via the World Wide Web, visit
> http://puck.nether.net/mailman/listinfo/juniper-nsp
> or, via email, send a message with subject or body 'help' to
> juniper-nsp-request at puck.nether.net
>
> You can reach the person managing the list at
> juniper-nsp-owner at puck.nether.net
>
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of juniper-nsp digest..."
>
>
> Today's Topics:
>
> 1. BGP load balance per packet (tchoel .)
> 2. How to determine last ifindex used ? (joe mcguckin)
> 3. Re: How to determine last ifindex used ? (Richard A Steenbergen)
> 4. Re: SMZ to BNC cables (joe lin)
>
>
> ----------------------------------------------------------------------
>
> Message: 1
> Date: Sat, 01 Nov 2003 21:42:56 +0000
> From: "tchoel ." <tchoel at hotmail.com>
> Subject: [j-nsp] BGP load balance per packet
> To: juniper-nsp at puck.nether.net
> Message-ID: <Law12-F41sgVy134IpL00002397 at hotmail.com>
> Content-Type: text/plain; format=flowed
>
> Hi,
>
> We've configured a JunOS 5.7R2.4 for BGP load-balancing.
> We've four links on our router all connected to same
> upstream autonomous system. Two of the links are connected
> to one upstream router and the other two is connected to an
> other upstream router. BGP sessions are established with
> loopback addresses and we've also configured static routes
> for each loop-back address. Import policy filters are setting
> the next-hop as the proper loopback address.
> [routing-options forwarding-table] is configured with a policy
> for load balance per-packet and bgp is configured with multipath.
> But we still see that traffic is forwarded using only one link
> of the each pair connected to upstream routers. I've checked the
> documentation but couldn't find a solution.Here's the configuration
> information;
>
> thanks,
> tchoel
>
> [routing-options static]
>
> route 192.168.0.1/32 next-hop [172.16.0.1 172.16.0.5] file://links A and B
> route 192.168.1.1/32 next-hop [172.16.1.1 172.16.1.5] file://links C and D
>
> [protocols bgp]
>
> group links-AB-toUp1 {
> type external;
> multihop {
> ttl 2;
> }
> local-address 10.0.0.1;
> import import4AB;
> export exportPolicy;
> peer-as 65513;
> multipath;
> neighbor 192.168.0.1;
> }
> group links-CD-toUp1 {
> type external;
> multihop {
> ttl 2;
> }
> local-address 10.0.0.1;
> import import4CD;
> export exportPolicy;
> peer-as 65513;
> multipath;
> neighbor 192.168.1.1;
> }
>
> [policy-options policy-statement import4AB]
>
> term first {
> from {
> ..
> things_to_reject;
> ..
> }
> then reject;
> }
>
> term second {
> from {
> protocol bgp;
> route-filter 0.0.0.0/0 upto /24;
> }
> then {
> next-hop 192.168.0.1;
> accept;
> }
> }
> term last {
> then reject;
> }
>
> [policy-options policy-statement import4CD]
>
> term first {
> from {
> ..
> things_to_reject;
> ..
> }
> then reject;
> }
>
> term second {
> from {
> protocol bgp;
> route-filter 0.0.0.0/0 upto /24;
> }
> then {
> next-hop 192.168.1.1;
> accept;
> }
> }
> term last {
> then reject;
> }
>
> [policy-options policy-statement equal_cost]
>
> then {
> load-balance per-packet;
> }
>
> [routing-options forwarding-table]
> export equal_cost;
>
> _________________________________________________________________
> MSN 8 helps eliminate e-mail viruses. Get 2 months FREE*.
> http://join.msn.com/?page=features/virus
>
>
>
> ------------------------------
>
> Message: 2
> Date: Sat, 01 Nov 2003 14:46:42 -0800
> From: joe mcguckin <joe at via.net>
> Subject: [j-nsp] How to determine last ifindex used ?
> To: "Juniper discussion thread (E-mail)" <juniper-nsp at puck.nether.net>
> Message-ID: <BBC977D2.46B11%joe at via.net>
> Content-Type: text/plain; charset="US-ASCII"
>
>
> Since ifindexes are not allocated contiguously, what's the best way to
> determine the maximum valid ifindex?
>
> Joe
>
>
>
> ------------------------------
>
> Message: 3
> Date: Sat, 1 Nov 2003 22:35:07 -0500
> From: Richard A Steenbergen <ras at e-gerbil.net>
> Subject: Re: [j-nsp] How to determine last ifindex used ?
> To: joe mcguckin <joe at via.net>
> Cc: "Juniper discussion thread \(E-mail\)"
> <juniper-nsp at puck.nether.net>
> Message-ID: <20031102033507.GO82121 at overlord.e-gerbil.net>
> Content-Type: text/plain; charset=us-ascii
>
> On Sat, Nov 01, 2003 at 02:46:42PM -0800, joe mcguckin wrote:
> >
> > Since ifindexes are not allocated contiguously, what's the best way to
> > determine the maximum valid ifindex?
>
> Are you sure that the maximum ifindex is what you're looking for, rather
> than the total count of ifindex's? You probably want the latter, since the
> former is fairly useless (especially if you're polling one of those
> not-a-juniper devices with weird fixed ifindex's of 50000 built in). The
> answer to the question you probably meant to ask is,
interfaces.ifNumber.0.
> You getnext up through your ifindex's until you hit the max count,
> skipping any unallocated ifindex's along the way.
>
> A smart design will poll the ifNumber.0 to check for changes, then run
> parallel get's of the ifName table. If any changes are detected, or if
> this is your first run, do a sequential getnext loop to map out / update
> the ifName's and ifIndex's, else do your normal speedy parallel polling.
> This lets you auto-detect interface changes (people who write pollers
> which need cfg updates or cfgmaker runs with every interface change need
> to be shot), without being piss slow on devices that aren't nearby.
>
> --
> Richard A Steenbergen <ras at e-gerbil.net> http://www.e-gerbil.net/ras
> GPG Key ID: 0xF8B12CBC (7535 7F59 8204 ED1F CC1C 53AF 4C41 5ECA F8B1 2CBC)
>
>
> ------------------------------
>
> Message: 4
> Date: Sat, 1 Nov 2003 22:51:35 -0800
> From: "joe lin" <jlin at doradosoftware.com>
> Subject: Re: [j-nsp] SMZ to BNC cables
> To: "Link King" <king at kinger.net>, <juniper-nsp at puck.nether.net>
> Message-ID: <00b301c3a10d$c1bef570$a20aa8c0 at oware.net>
> Content-Type: text/plain; charset="iso-8859-1"
>
> the connector is a 75ohm posilink SMB type connector. maybe you can crimp
> some of your own?
>
> ----- Original Message -----
> From: "Link King" <king at kinger.net>
> To: <juniper-nsp at puck.nether.net>
> Sent: Friday, October 31, 2003 8:47 AM
> Subject: [j-nsp] SMZ to BNC cables
>
>
> >
> > Anyone know of an alternative to Juniper's CBL-SMZ-BNC-MS cable for
their
> > 4 port DS3 cards? We actually only need one port so 2 cables would
> > suffice. Any generic cable vendors carry these?
> >
> > --
> > Link King
> > king at kinger.net
> >
> > _______________________________________________
> > juniper-nsp mailing list juniper-nsp at puck.nether.net
> > http://puck.nether.net/mailman/listinfo/juniper-nsp
> >
> >
>
>
>
>
> ------------------------------
>
> _______________________________________________
> juniper-nsp mailing list
> juniper-nsp at puck.nether.net
> http://puck.nether.net/mailman/listinfo/juniper-nsp
>
>
> End of juniper-nsp Digest, Vol 12, Issue 2
> ******************************************
>
More information about the juniper-nsp
mailing list