John Arnold (GNS | 26 Feb 01:13 2015
Picon

[quagga-users 13926] Re: subnet /31

Hi,

 

Is configuring Ethernet interfaces with /31 via zebra bugged?  Broadcast address gets configured as 0.0.0.0 and pings to the far side don’t work.  I’ve reviewed RFC3021 and I think the broadcast address is supposed to be set to link local broadcast,  255.255.255.255.

 

>> Apply /31 via quagga:

quagga-router# conf t

quagga-router(config)# int e00-1

quagga-router(config-if)# ip address 50.0.0.0/31

quagga-router(config-if)# end

quagga-router# write mem

Building Configuration...

Integrated configuration saved to /etc/quagga/Quagga.conf

[OK]

 

root <at> r1:/home/amazon# ifconfig e00-1

e00-1     Link encap:Ethernet  HWaddr ec:f4:bb:fc:21:74

          inet addr:50.0.0.0  Bcast:0.0.0.0  Mask:255.255.255.254

          inet6 addr: fe80::eef4:bbff:fefc:2174/64 Scope:Link

          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1

          RX packets:0 errors:0 dropped:0 overruns:0 frame:0

          TX packets:25599 errors:0 dropped:0 overruns:0 carrier:0

          collisions:0 txqueuelen:500

          RX bytes:0 (0.0 B)  TX bytes:4786165 (4.7 MB)

 

 

Applying ip/subnet via ifconfig works (pings work) yet gives a weird broadcast address:

 

>> apply /31 via ifconfig:

 

root <at> amazon:/home/amazon# ifconfig e00-1 50.0.0.0/31

root <at> amazon:/home/amazon# ifconfig e00-1

e00-1     Link encap:Ethernet  HWaddr ec:f4:bb:fc:21:74

          inet addr:50.0.0.0  Bcast:50.255.255.255  Mask:255.255.255.254

          inet6 addr: fe80::eef4:bbff:fefc:2174/64 Scope:Link

          UP BROADCAST MULTICAST  MTU:1500  Metric:1

          RX packets:0 errors:0 dropped:0 overruns:0 frame:0

          TX packets:25605 errors:0 dropped:0 overruns:0 carrier:0

          collisions:0 txqueuelen:500

          RX bytes:0 (0.0 B)  TX bytes:4787287 (4.7 MB)

 

And anecdotally I found some references to RFC3021 broadcast addresses being miscalculated in  ifupdown code, so perhaps ifconfig is bugged too.

 

Using the newer libraries I see the same result as quagga (broadcast 0.0.0.0):

 

>> add /31 via iproute2 tools

 

root <at> amazon:/home/amazon# ip addr add 50.0.0.0/31 dev e00-1

root <at> amazon:/home/amazon# ip addr show dev e00-1

8: e00-1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP group default qlen 500

    link/ether ec:f4:bb:fc:21:74 brd ff:ff:ff:ff:ff:ff

    inet 50.0.0.0/31 scope global e00-1

       valid_lft forever preferred_lft forever

    inet6 fe80::eef4:bbff:fefc:2174/64 scope link

       valid_lft forever preferred_lft forever

root <at> amazon:/home/amazon# ifconfig e00-1

e00-1     Link encap:Ethernet  HWaddr ec:f4:bb:fc:21:74

          inet addr:50.0.0.0  Bcast:0.0.0.0  Mask:255.255.255.254

          inet6 addr: fe80::eef4:bbff:fefc:2174/64 Scope:Link

          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1

          RX packets:0 errors:0 dropped:0 overruns:0 frame:0

          TX packets:25613 errors:0 dropped:0 overruns:0 carrier:0

          collisions:0 txqueuelen:500

          RX bytes:0 (0.0 B)  TX bytes:4788783 (4.7 MB)

 

Anybody know what’s going on here?

 

Thanks,

John

_______________________________________________
Quagga-users mailing list
Quagga-users@...
https://lists.quagga.net/mailman/listinfo/quagga-users
Julio Tejedor | 10 Feb 12:37 2015
Picon

[quagga-users 13921] Doubt. Is it possible to use OSPF NSF or OSPF GraceFul Restart?

Hello,

We are trying to include OSPF NSF (OSPFv2 or OSPF GraceFul Restart
(OSPFv3) in our quaga equipment. We have looked for this in the
documentation but nothing has been found. Could you please give us any
clue about how to configure this functionality?

KR
Julio
Pablo Mellado | 4 Feb 15:00 2015
Picon

[quagga-users 13919] OSPF6 daemon aborting

Hi there,

Yesterday I tried to set up ospf6 daemon and it was aborting:

2015/02/03 13:55:24 OSPF6: OSPF6d (Quagga-0.99.22.4 ospf6d-0.9.7r) 
starts: vty <at> 2606
2015/02/03 13:55:24 OSPF6: Neighbor state change [ipv4 deleted]%eth1: 
[Down]->[Init]
2015/02/03 13:55:24 OSPF6: Neighbor state change [ipv4 deleted]%eth1: 
[Init]->[Twoway]
2015/02/03 13:55:24 OSPF6: Neighbor state change [ipv4 deleted]%eth1: 
[Twoway]->[ExStart]
2015/02/03 13:55:24 OSPF6: Neighbor state change [ipv4 deleted]%eth1: 
[ExStart]->[ExChange]
2015/02/03 13:55:24 OSPF6: Neighbor state change [ipv4 deleted]%eth1: 
[ExChange]->[Loading]
2015/02/03 13:55:24 OSPF6: Neighbor state change [ipv4 deleted]%eth1: 
[Loading]->[Full]
2015/02/03 13:55:29 OSPF6: PANIC !!
2015/02/03 13:55:29 OSPF6: Something has gone wrong with 
ospf6_route_table[0x7f3501baf5c0]
2015/02/03 13:55:29 OSPF6: table count = 4, real number = 3
2015/02/03 13:55:29 OSPF6: DUMP START
2015/02/03 13:55:29 OSPF6: (nil)<-[0x7f3501bca690]->0x7f3501be03d0 
:[ipv6 deleted] /126
2015/02/03 13:55:29 OSPF6: 
0x7f3501bca690<-[0x7f3501be03d0]->0x7f3501bcdf40 : 2a00:[ipv6 deleted] /126
2015/02/03 13:55:29 OSPF6: 0x7f3501be03d0<-[0x7f3501bcdf40]->(nil) : 
2a00:[ipv6 network deleted] /64
2015/02/03 13:55:29 OSPF6: DUMP END
2015/02/03 13:55:29 OSPF6: Assertion `link_error == 0 && num == 
table->count' failed in file ospf6_route.c, line 348, function 
route_table_assert
2015/02/03 13:55:29 OSPF6: Backtrace for 11 stack frames:
2015/02/03 13:55:29 OSPF6: [bt 0] 
/usr/lib/libzebra.so.0(zlog_backtrace+0x26) [0x7f350083c673]
2015/02/03 13:55:29 OSPF6: [bt 1] 
/usr/lib/libzebra.so.0(_zlog_assert_failed+0xa3) [0x7f350083cd76]
2015/02/03 13:55:29 OSPF6: [bt 2] /usr/lib/quagga/ospf6d(+0x1ef25) 
[0x7f3500caaf25]
2015/02/03 13:55:29 OSPF6: [bt 3] 
/usr/lib/quagga/ospf6d(ospf6_route_add+0x742) [0x7f3500cab834]
2015/02/03 13:55:29 OSPF6: [bt 4] 
/usr/lib/quagga/ospf6d(ospf6_intra_prefix_lsa_add+0x2e5) [0x7f3500caf281]
2015/02/03 13:55:29 OSPF6: [bt 5] 
/usr/lib/quagga/ospf6d(ospf6_intra_route_calculation+0x83) [0x7f3500caf5d9]
2015/02/03 13:55:29 OSPF6: [bt 6] /usr/lib/quagga/ospf6d(+0x259ab) 
[0x7f3500cb19ab]
2015/02/03 13:55:29 OSPF6: [bt 7] 
/usr/lib/libzebra.so.0(thread_call+0x8f) [0x7f3500831910]
2015/02/03 13:55:29 OSPF6: [bt 8] /usr/lib/quagga/ospf6d(main+0x39e) 
[0x7f3500c9ac1e]
2015/02/03 13:55:29 OSPF6: [bt 9] 
/lib/x86_64-linux-gnu/libc.so.6(__libc_start_main+0xfd) [0x7f34ffe5dead]
2015/02/03 13:55:29 OSPF6: [bt 10] /usr/lib/quagga/ospf6d(+0xec85) 
[0x7f3500c9ac85]
Aborted

Have you seen this behavior before?

BTW, I have remove IP addresses to not compromise security.

Thanks for reading this,

   Pablo

--

-- 

Pablo Mellado
Software Engineer

Instituto de Radioastronomia Milimetrica
Avenida Divina Pastora 7, Local 20
E 18012 Granada, España
Telefono: +34 958 805 456 Fax: +34 958 222363

_______________________________________________
Quagga-users mailing list
Quagga-users <at> lists.quagga.net
https://lists.quagga.net/mailman/listinfo/quagga-users
Balaji G | 30 Jan 13:48 2015
Picon

[quagga-users 13912] Re: [quagga-dev 11998] Difference between Quagga and cisco, junos routers



On Fri, Jan 30, 2015 at 5:58 PM, Rajesh B <rajesh.1982.bglr <at> gmail.com> wrote:
Thanks Balaji.. Very nice explanation.

Where this control plane software is used. For what reasons this software is built for, Can this software will be used for high traffic/commercial purpose.
Control plane is built for populating the NP in case of routers having it and in case of Linux machines for populating the kernel's FIB. Control plane protocols use best path algorithms to choose the best path and the same gets installed in the kernel/NP. These operations are carried by routing protocols running in user space which is also called as control plane. Traffic is again handled by the FIB as the FIB gets populated (In Unicast) even before the traffic comes into the box and based on the route, packet gets forwarded out 

 


On Fri, Jan 30, 2015 at 5:41 PM, Balaji G <balajig81 <at> gmail.com> wrote:


On Fri, Jan 30, 2015 at 5:36 PM, Rajesh B <rajesh.1982.bglr-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> wrote:
So the calculation of routes by using different algo's is similar to other networking vendors except we use Linux kernel for forwarding.
If we have higher end specification of linux machines, then can we easily replace the cisco/juniper routers with this linux router?
 
One more question, I am not aware about ASIC/Network processor. Does Linux kernel is capable to handle(forward) traffic similar to ASIC/Network processor.
When packets are forwarded using NP it means they are hardware forwarded and generally they have TCAMs where the lookups are done and its faster. When packets are forwarded using the OS it means the packets are software forwarded and cannot match the speed of the H/w. Its the forwarding engine that makes the router more efficient as the control plane stuff does not come into picture during forwarding  


On Fri, Jan 30, 2015 at 4:59 PM, <sthaug <at> nethelp.no> wrote:
> The efficiency of algorithm is implemented in quagga. So does this hamper
> the forwarding procedure to delay?
> Correct me if my understanding is wrong.

The Quagga implementation runs SPF and similar algorithms. As long as
Quagga calculates the same SPF tree, the forwarding delay is *not*
related to Quagga (but to the kernel).

Steinar Haug, Nethelp consulting, sthaug-5U4UkxQHZ+huMpJDpNschA@public.gmane.org


_______________________________________________
Quagga-users mailing list
Quagga-users-UOy77sIEA+cAd7ICUelF/Q@public.gmane.org
https://lists.quagga.net/mailman/listinfo/quagga-users




_______________________________________________
Quagga-users mailing list
Quagga-users@...
https://lists.quagga.net/mailman/listinfo/quagga-users
Jim Carroll | 30 Jan 13:14 2015

[quagga-users 13911] Quagga on windows

I’m sure this has been discussed, but I can’t find the previous thread(s) in the list archive.  Has there been any success trying to port quagga to Windows?  Our operation is using more W2k8+ which no longer runs OSPF. We’ve experimented with xorp, but it’s no where near as good as quagga which we’ve adopted as our organization’s standard routing s/w.

 

We’ve started exploring trying to port quagga to windows using Cygwin, and wanted to know if this had already been attempted and if people might share their experiences.

 

Jim

_______________________________________________
Quagga-users mailing list
Quagga-users@...
https://lists.quagga.net/mailman/listinfo/quagga-users
Zhenyu Zhou | 28 Jan 00:08 2015

[quagga-users 13903] How to combine Quagga with Floodlight?

Hi all,

I'm looking for a L3 routing module for floodlight and someone told me 
Quagga may be a solution. But I didn't find how to combine Quagga with 
floodlight. Any idea about this issue? Is it possible to regard Quagga 
as a floodlight module?

Many thanks!

--

-- 
Best Regards!

Zhou, Zhenyu
Daniel STICKNEY | 21 Jan 17:03 2015

[quagga-users 13892] Route-map not found with OSPF6

Hello all,

I've run into a problem with OSPF6 and redistributing routes with a
route-map using Quagga 0.99.22.4 on Debian 7.

I've done searches in the Quagga Bugzilla but I didn't find others
reporting route-map "not found" messages.

Starting from a functioning base configuration (OSPF6 working with no
redistribution), here is how I reproduced the problem :

conf t
ipv6 prefix-list PL_OPTLN6 seq 20 permit 2a00:18a8:400::/62
ipv6 prefix-list PL_OPTLN6 seq 100 deny any
route-map OPT-OSPF6-RDIST permit 10
  match  ipv6  address  prefix-list  PL_OPTLN6
  exit
router ospf6
  redistribute  connected  route-map  OPT-OSPF6-RDIST
  redistribute  kernel  route-map  OPT-OSPF6-RDIST
  redistribute  static  route-map  OPT-OSPF6-RDIST
  exit
ipv6  route  2a00:18a8:400::/62 lo
exit
wr mem

At this point the redistribution of this static route works, and I see
the static route in the routing tables of the other OSPF6 routers on the
network.

I then restart Quagga
 /etc/init.d/quagga restart

I then see this message in Quagga.log
2015/01/21 16:30:32 OSPF6: route-map "OPT-OSPF6-RDIST" not found,
suppress redistributing

OSPF6d establishes adjacencies with the other OSPF6 routers but
redistributes nothing, as expected after seeing this log message.

I've tried different route-map names, such as RDIST, with no special
characters, and I get the exact same results.

To check that OSPF6d can see the route-map I connect to OSPF6d via
telnet and execute the following command, which confirms that the
route-map does exist and is seen by OSPF6d.

# show route-map OPT-OSPF6-RDIST
OSPF6:
route-map OPT-OSPF6-RDIST, permit, sequence 10
  Match clauses:
    ipv6 address prefix-list PL_OPTLN6
  Set clauses:
  Call clause:
  Action:
    Exit routemap

I am at a loss for a solution to this problem. Does anyone have any
suggestions?

Thanks for your time,
Daniel
Indo Php | 19 Jan 06:44 2015
Picon

[quagga-users 13888] Quagga not updated.

Hi

I'm using Quagga for my BGP router. I've strange issue with Quagga

Sample :
netstat -rn | grep 185.67.164
It shows nothing

But sh ip bgp 185.67.164 shows
BGP routing table entry for 185.67.0.0/24
Paths: (1 available, best #1, table Default-IP-Routing-Table)
  Not advertised to any peer
  asn1 asn2 asn3 asn4 50673
    x.x.x.x from y.y.y.y (y.y.y.y)
      Origin IGP, metric 0, localpref 250, valid, internal, best
      Last update: Sun Jan 18 14:42:05 2015

sh ip route 185.67.164 shows
Routing entry for 185.67.0.0/24
  Known via "bgp", distance 200, metric 0, best
  Last update 21:56:16 ago
    x.x.x.x (recursive)
  *   z.z.z.z, via vlan20

Also when doing debug, I got logs like below
2015/01/19 12:39:32 ZEBRA: rtm_read: RTM_DELETE 185.67.0.0/24: desync: RR is still in RIB, while already not in FIB

Can somebody tell me what's wrong? 

Note : I'm using OpenBSD 5.3 with Quagga version 0.99.23.1

Thank you
 
_______________________________________________
Quagga-users mailing list
Quagga-users@...
https://lists.quagga.net/mailman/listinfo/quagga-users
Timo Schöler | 17 Jan 12:47 2015
Picon

[quagga-users 13885] BCP38 on a Linux/Quagga environment

Hi list,

just a short question: What's the most effective way to implement BCP38 
on a bunch of BGP routers based on Linux/Quagga -- using Linux' on-board 
techniques, configuring Quagga or a mix of both?

Thanks,

Timo
Jiann-Ming Su | 16 Jan 21:10 2015
Picon

[quagga-users 13883] OSPF Metric for Loopback?

How should quagga/ospfd be configured such that the ip ospf cost configured for a loopback is added to the metric cost when the host route gets redistributed?

Sample ospfd config:

interface lo:0
 ip ospf cost 100

The router that the quagga server peers with has ip ospf cost 500.  The metric for the anycast associated with lo:0 shows the 500 added, but not the additional 100 listed in the ospfd config.  Thanks for any tips.

_______________________________________________
Quagga-users mailing list
Quagga-users@...
https://lists.quagga.net/mailman/listinfo/quagga-users
Daniel Paez Sanchez | 16 Jan 21:11 2015
Picon

[quagga-users 13884] RIPng parameters in quagga

How can I set priority to an interface in the file ripng.conf.samples ?

--
Daniel Páez

_______________________________________________
Quagga-users mailing list
Quagga-users@...
https://lists.quagga.net/mailman/listinfo/quagga-users

Gmane