headers
James Andrewartha | 1 Feb 02:52
Picon
Favicon

[quagga-users 12704] BGP routes becoming inactive after peer drop

Hi,

We have two BGP sessions, one providing a full feed, and another just
providing a default route. Whenever the full feed peer drops, all BGP
routes are marked as inactive and don't appear in the kernel's routing
table. Even if the full feed comes back it's still inactive. The only
way to fix it is to restart quagga.

Here's an excerpt of our config, the full feed comes in from
115.70.192.5 over vlan1025:

interface vlan1025
 ip address 115.70.192.6/30
 ipv6 nd suppress-ra
!
interface vlan1026
 ip address 203.24.54.150/30
 ipv6 nd suppress-ra
!
router bgp 45797
 bgp router-id 203.135.184.1
 bgp log-neighbor-changes
 network 203.135.184.0/22
 network 203.135.184.0/23
 network 203.135.186.0/23
 neighbor 115.70.192.5 remote-as 10143
 neighbor 115.70.192.5 description Exetel
 neighbor 115.70.192.5 weight 2000
 neighbor 115.70.192.5 prefix-list split23 out
 neighbor 150.101.172.174 remote-as 4739
(Continue reading)

Permalink | Reply | 6 comments |
headers
Denis Ovsienko | 1 Feb 07:31
Picon
Favicon
Gravatar

[quagga-users 12705] Re: BGP routes becoming inactive after peer drop

01.02.2012, 05:52, "James Andrewartha" <jandrewartha@...>:
> Hi,
>
> We have two BGP sessions, one providing a full feed, and another just
> providing a default route. Whenever the full feed peer drops, all BGP
> routes are marked as inactive and don't appear in the kernel's routing
> table. Even if the full feed comes back it's still inactive. The only
> way to fix it is to restart quagga.

Does it reproduce on the latest release? What would "show ip bgp" output after the full view feed is down?

--

-- 
    Denis Ovsienko
Permalink | Reply | 5 comments |
headers
Subbaiah Venkata | 1 Feb 10:52
Picon
Favicon

[quagga-users 12706] Re: [quagga-dev 8742] Google changes to Quagga now available on code.google.com

Hi All,

New ISIS branch is pushed to google public repository. 
http://code.google.com/p/google-quagga/source/browse/?name=is-is

This branch has support for new features and bug fixes.
Please let me know if you have any questions.

Here are the steps to enable ISIS:
$ git clone https://code.google.com/p/google-quagga/
$ cd google-quagga
$ git checkout is-is
$ ./bootstrap.sh
$ ./configure --enable-vtysh --enable-isisd --enable-multipath=0 
$ make
$ make install

Special thanks to Josh for taking care of all the logistics. 

Thank You,
--
Subbaiah Venkata
Technical Leader and Manager
Google Inc.
1600 Amphitheatre Parkway
Mountain View, CA 94043


On Thu, Jul 21, 2011 at 2:34 PM, Josh Bailey <joshb-hpIqsD4AKlfQT0dZR+AlfA@public.gmane.org> wrote:

Hi all;

I'd like to announce a newly set up git repo at code.google.com, where I will be releasing changes by Google to Quagga - both features and fixes.

The first branch I've made available implements BGP multipath (by our own Barry Friedman).

I will be announcing/adding further features in the coming weeks, and we have plans in the works with our friends at the ISC to provide SQA coverage for all these changes.

Here's a quick cheat sheet on how to obtain and build the BGP multipath feature:

$ git clone https://code.google.com/p/google-quagga/ bgp-multipath
$ cd bgp-multipath
$ git checkout bgp-multipath
$ ./bootstrap.sh
$ ./configure
$ make

Any questions, please let me know!

Thanks,

--
Josh Bailey
_______________________________________________
Quagga-dev mailing list
Quagga-dev <at> lists.quagga.net
http://lists.quagga.net/mailman/listinfo/quagga-dev
 

_______________________________________________
Quagga-users mailing list
Quagga-users@...
http://lists.quagga.net/mailman/listinfo/quagga-users
Permalink | Reply | 9 comments |
headers
narender | 2 Feb 07:35
Picon

[quagga-users 12707] Re: TODO list for OSPFv3 in the preference order

Tore Anderson <tore.anderson@...> writes:

> Hi,
> 
> Please feel free to have a look at this one:
> 
> https://bugzilla.quagga.net/show_bug.cgi?id=549
> 
> 
> 
> Best regards,

I have seen in the mailing lists that multiple areas feature is not implemented 
yet. But when i test two areas (Area ID: 0.0.0.0 and Area ID2: 0.0.0.1)
configured on a router, I could see the adjacencies properly formed and LSDB 
properly exchanged. 

Is multi areas supported in ospf6d or i failed to notice something. 
Has any 1 tried using multi areas configured in ospf6d. What is exactly
missing here. 

Thanks
Permalink | Reply | 1 comment |
headers
Tore Anderson | 2 Feb 11:23

[quagga-users 12708] Re: TODO list for OSPFv3 in the preference order

* narender

> I have seen in the mailing lists that multiple areas feature is not implemented 
> yet. But when i test two areas (Area ID: 0.0.0.0 and Area ID2: 0.0.0.1)
> configured on a router, I could see the adjacencies properly formed and LSDB 
> properly exchanged. 
> 
> Is multi areas supported in ospf6d or i failed to notice something. 
> Has any 1 tried using multi areas configured in ospf6d. What is exactly
> missing here. 

I'm not using Quagga as an ABR, so I don't know about that.

What I would like is to be able to have OSPFv3 running on all interfaces
in passive/stub mode by default - with the possiblity to exempt certain
interfaces (typically the uplink interface) from that. The idea is that
if I bring up another interface on the router, that link will be
advertised in OSPFv3 with no configuration changes required.

With OSPFv2 I can easily do this:

  router ospf
    passive-interface default <--- by default use passive mode on all
                                   interfaces
    no passive-interface eth0 <--- this is the uplink interface, so
                                   don't use passive here
    network 0.0.0.0/0 area 0.0.0.0 <-- run ospf on all interfaces with
                                       an IPv4 address

If I bring up a new interface ("vlan123", let's say) and add an IPv4
address to it, it would get advertised in OSPFv2 with *no* ospfd
configuration changes required. This can't be done in OSPFv3, every new
interface requires two lines of explicit configuration:

  interface vlan123
    ipv6 ospf6 passive <--- must be set explicitly per interface
  router ospf6
    interface vlan123 area 0.0.0.0 <--- each interface must be
                                        explicitly added here too

So for a typical data centre access router with one or two uplink
interfaces and hundreds of dynamic VLAN interfaces, I never have to
touch the ospfd config. The ospf6d config, however, I need to change
every time I make changes to the VLANs themselves. I really wish I
didn't have to do that.

Oh and another ting - NSSA support in OSPFv3 would be great to have...

Best regards,
--

-- 
Tore Anderson
Redpill Linpro AS - http://www.redpill-linpro.com
Permalink | Reply | 0 comments |
headers
Steve Clark | 2 Feb 12:54

[quagga-users 12709] OSPF over IPSEC without GRE

Hi all,

Cisco has come up with a way to do OSPF over IPSEC without using GRE. Can this same
thing be accomplished with quagga and Linux?

http://www.cisco.com/en/US/products/hw/vpndevc/ps2030/products_configuration_example09186a00804acfea.shtml

Thanks,

--
Stephen Clark
NetWolves
Director of Technology
Phone: 813-579-3200
Fax: 813-882-0209
Email: steve.clark-HKs6b5iW9l2akBO8gow8eQ@public.gmane.org
http://www.netwolves.com
_______________________________________________
Quagga-users mailing list
Quagga-users@...
http://lists.quagga.net/mailman/listinfo/quagga-users
Permalink | Reply | 1 comment |
headers
Ryan Whelan | 2 Feb 15:31
Picon
Gravatar

[quagga-users 12710] Re: OSPF over IPSEC without GRE

I've not tried, but it seems to me if you used klips (vs netkey) you'd have a virtual IPSec network interface you could configure Quagga to use. 


On Thu, Feb 2, 2012 at 6:54 AM, Steve Clark <sclark-HKs6b5iW9l2akBO8gow8eQ@public.gmane.org> wrote:
Hi all,

Cisco has come up with a way to do OSPF over IPSEC without using GRE. Can this same
thing be accomplished with quagga and Linux?

http://www.cisco.com/en/US/products/hw/vpndevc/ps2030/products_configuration_example09186a00804acfea.shtml

Thanks,

--
Stephen Clark
NetWolves
Director of Technology
Phone: 813-579-3200
Fax: 813-882-0209
Email: steve.clark-HKs6b5iW9l2akBO8gow8eQ@public.gmane.org
http://www.netwolves.com

_______________________________________________
Quagga-users mailing list
Quagga-users-UOy77sIEA+f4/q2+jIuWiA@public.gmane.orgnet
http://lists.quagga.net/mailman/listinfo/quagga-users


_______________________________________________
Quagga-users mailing list
Quagga-users@...
http://lists.quagga.net/mailman/listinfo/quagga-users
Permalink | Reply | 0 comments |
headers
Sanjeev Gupta | 6 Feb 07:30

[quagga-users 12711] BPG Proxy objects

Hi,

This is not a quagga question, but since I use quagga, this is the only BGP community I frequent.

Over the years, my ASN (37989) has been included by various transit providers in proxy objects.  Some of these are no longer providing transit for me.

Does this hurt?
Do I need to clean these?

Thank you,
--
Sanjeev Gupta
+65 98551208

_______________________________________________
Quagga-users mailing list
Quagga-users@...
http://lists.quagga.net/mailman/listinfo/quagga-users
Permalink | Reply | 1 comment |
headers
Nick Hilliard | 6 Feb 12:53
Picon

[quagga-users 12712] Re: BPG Proxy objects

On 06/02/2012 06:30, Sanjeev Gupta wrote:
> This is not a quagga question, but since I use quagga, this is the only BGP
> community I frequent.
> 
> Over the years, my ASN (37989) has been included by various transit
> providers in proxy objects.  Some of these are no longer providing transit
> for me.
> 
> Does this hurt?
> Do I need to clean these?

This isn't really related to quagga in any way, but:

It will hurt you if you ever interconnect with an organisation which uses
IRRDB objects for filtering your prefixes.  Best thing to do is to maintain
your own AS object in the APNIC IRRDB, and to get your old transit
providers to delete any proxy objects they may have registered.

Nick
Permalink | Reply | 0 comments |
headers
Juliusz Chroboczek | 7 Feb 22:46
Picon

[quagga-users 12713] Multiple routing tables (Linux policy routing)

Hi,

My router is running quagga in a non-default table (table 8), since I'm
using different default routes for locally originated packets and for
forwarded packets (which go through a NAT box):

  $ grep table /usr/local/quagga/etc/zebra.conf
  table 8
  $ ip rule show
  0:      from all lookup local 
  100:    from 192.168.4.0/24 lookup 8 
  101:    from all to 192.168.4.0/24 lookup 8 
  32766:  from all lookup main 
  32767:  from all lookup default 

I'd like there to be two distinct default routes that are active at the
same time, one to the (quagga discovered) NAT box in table 8, one to the
(statically configured) router in table main.  Unfortunately, quagga
doesn't seem to install the route in table 8:

  $ ip route show 0.0.0.0/0 table main
  default via A.B.C.D dev eth0 
  $ ip route show 0.0.0.0/0 table 8
  $

Quagga is obviously receiving the default route to the NAT box:

  # show ip route 0.0.0.0/0
  Routing entry for 0.0.0.0/0
    Known via "babel", distance 95, metric 0
      E.F.G.H inactive
      directly connected, coloquinte

  Routing entry for 0.0.0.0/0
    Known via "kernel", distance 0, metric 0, best
    * A.B.C.D, via eth0

Apparently, quagga is ignoring the table number when checking for
duplicates, and refusing to install a duplicate default route, although
the two routes are in different tables.  Is there any way to force it to
install the route?

Is that a bug?

-- Juliusz
Permalink | Reply | 2 comments |

Gmane