Fine then! I was suspicious about some changes in thread-next (where this
problem can be actually reproduced with a good probability) and now I know that
I should assess the whole algorithm instead.
Thanks a lot!
Maria
On October 7, 2022 11:29:09 AM UTC, Mihai wrote:
>Hi Maria,
>
>Yes you are
Hello!
As there are some merges done, here is a short update.
>3G. the current route attribute implementation is a two-layer mess originating
>in version 1 where everything was an IP route, let's squash these into one
>layer [IN PROGRESS]
Done (at least for now), will also comment the
Hello!
I'm curious as well and I'd like to ask you also, if you could, to send me
(privately) a binary and a corefile of such a case you are writing about. I'll
inspect the memory and tell exactly how much you'd be affected by the u32->u64
change.
Thanks a lot!
Maria
On March 28, 2022
Hello!
Do you have any log? Are you able to replicate such a behavior consistently? If
so, could you please share an exact configuration with us to put it into our
testbed?
You can also enable "debug protocols all;" in your conf file. This produces a
s***load of logs, yet it should yield
Hello!
You can try capturing the packets in wire by tcpdump. It should tell you what
exactly is being sent.
Maria
On August 24, 2021 8:17:26 PM UTC, Noris Rogers wrote:
>Hello,
>
>I am seeking to setup BIRD as a flowspec manager to send messages to FRR
>acting as a flowspec client. However,
Hello!
I'd try:
* "show route noexport all" to see what may be the reason to drop the route
* add some logging to the filter to see why it is dropped
BTW, what version are you using? As far as I remember, the BGP attributes are
behaving differently on export in 1.6 and 2.0
Maria
On June 6,
Hello!
If you have one router connecting these networks, this is just a matter of
static routing, no need for BIRD. Otherwise, you may probably want to setup
Babel or OSPF.
Maria
On April 27, 2021 12:04:30 AM GMT+02:00, Myron wrote:
>I am trying to learn how to set-up a router and maintain a
Hello!
Yes, it is going to happen. The order should be (roughly) as follows:
0) basic multithreading support (already implemented)
1) structural changes in route export enabling table access from multiple
threads
2) conversion of 'show route' to the new route export code
3) json route
of when these changes might get merged into master?
>
>Thanks!
>Trisha
>--
>
>*Trisha Biswas* | Sr. Software Engineer, Network Systems
>fastly.com | @fastly <https://twitter.com/fastly> | LinkedIn
><http://www.linkedin.com/company/fastly>
>
>
>On S
Hello!
According to our internal decisions in this week, we may be able to implement
some aggregations in one or two months, depending on how fast the needed
refactoring happens. Anyway, there is a multithreaded preview release in the
queue before this.
Maria
On February 18, 2021 1:47:25 PM
Hello!
This is currently not possible. There are some thoughts how to make it
possible, yet never finished and implemented.
Maria
On January 22, 2021 9:26:59 PM GMT+01:00, David Johnson wrote:
>Hi,
>
>I'm trying to see if bird can accomplish the following and so far I
>can't
>seem to find a way
Hello!
This means that you were refusing messages from the list for some amount of
time. You may want to check your mailserver.
Maria
On November 22, 2020 7:38:10 PM GMT+01:00, Jason Kopacko
wrote:
>What is this about
>
>I haven't sent anything to this list in a while.
>
Hello!
Just a quick reply from my phone. When adding these checks, I also wanted to
zero out slab objects but then I realized that these objects should be
initialized anyway after allocation and in most cases all of these would be
rewritten twice.
The preferred way of using slab objects is
Hello!
Would it be feasible for you to have a special route attribute to be set in
filters that would control the nexthop weight? That seems to be probably a most
simple solution.
Maria
On November 19, 2020 5:42:38 PM GMT+01:00, Nigel Kukard
wrote:
>On 11/19/20 4:06 PM, Alexander Demenshin
Hello!
Not for now. There is some extension of the filter language pending (not
implemented yet) which would allow for protocol attributes queries. For now,
you can make your filter a function and use different arguments in appropriate
calls.
Maria
On November 7, 2020 2:34:50 PM GMT+01:00,
. But from the BGP collectors perspective,
>we observe that around 50% of all announcements fall into the category
>(i) and (ii), in March 2020.
>
>Thanks for your feedback. It's very insightful!
>
>best regards
>Thomas
>
>On 10/17/20 10:39 AM, Maria Matějka wrote:
>
Hello!
If I read your paper correctly, I assume that this is just a feature of the BGP
itself. To make it clear: I read that in some cases when a link inside one AS
goes down, a route is reannounced with the same AS path but different
communities because of e.g. some internal tagging which you
with implementing Linux
nexthops.
Thank you for pointing it out, this is a very useful note from you.
Maria
On October 3, 2020 3:28:41 PM GMT+02:00, Marek Zarychta
wrote:
>On 23.09.2020 07:10, Maria Matějka wrote:
>> Hello!
>>
>> Yes, we definitely want to support it, yet last time whe
Hello!
It is quite a difficult task to dump the current config. Somewhere on the
roadmap, there is possibly a change of internal structures that would allow for
more versatile config, yet it is neither simple nor quickly done.
So it is currently not possible, sorry.
Maria
On September 29, 2020
Hello!
Yes, we definitely want to support it, yet last time when we were talking about
it, it was still unstable. It is good news that it really has come in, we'll
study it and most likely also implement it.
Thank you for pointing this out.
Maria
On September 22, 2020 11:08:02 PM GMT+02:00,
Hello!
This list is generally the right place to post your patches.
Thank you for your contribution.
Maria
On September 11, 2020 10:04:27 PM GMT+02:00, "Gehrkens.IT GmbH | Heiko Wundram"
wrote:
>Hey all,
>
>I’ve played around with bird a bit (which we use as a routing daemon,
>mainly for OSPF
Hello!
Just to be sure, does it work via iptools command, like ip route add…?
Thanks
Maria
On August 19, 2020 5:10:18 PM GMT+02:00, Mikhail Petrov
wrote:
>Hello!
>
>
>
>I have a problem with bird, static routes and multipath.
>
>My bird config is https://pastebin.com/gLkFP6rK
>
>And all work
Hello!
The error message tells you that you are passing something strange to the
condition on line 360. What do you have on line 360?
Maria
On August 13, 2020 4:46:12 PM GMT+02:00, "Skyler Mäntysaari"
wrote:
>Hi there,
>
>I'm using the template from
Hello!
BIRD reads the routes from kernel anyway, the import clause applies only on
learned external routes. Do you have any export filter on that kernel protocol?
Don't you have any suspicious warnings in log?
Maria
On August 2, 2020 7:08:11 PM GMT+02:00, Alarig Le Lay
wrote:
>Hi Sasha,
>
>On
Hello!
This seems to be useful for me also. Will implement after returning from
vacation unless anybody else is faster.
Maria
On July 3, 2020 9:09:23 PM GMT+02:00, Ross Tajvar wrote:
>I agree, this would be nice to have as a configurable option.
>
>On Fri, Jul 3, 2020, 4:33 AM Tim Small wrote:
Hello!
Yes, we're dropping support for such an old GCC, mostly because we don't have
enough resources (time) to manage those old systems for testing. Other problems
are with lack of support of some C11 features. I should also say that the
upcoming multithreaded versions will also need a
Hello,
Prefix aggregation is currently not supported. It's quite a lot of work to do
it properly.
As a workaround, you may export these routes to an auxiliary kernel table, run
an external script to maintain the aggregated route and learn the result back
to BIRD. Or you can simply use the
Hello!
There is a fundamental principle in BIRD that one BIRD table may be synced to
at most one kernel table. Use BIRD pipe pseudoprotocol to duplicate the routes
to more BIRD tables.
This concept may be changed in future, yet not in near future.
Maria
On June 1, 2020 7:58:39 PM GMT+02:00,
Hello!
We know about that, yet we have postponed it at least after some changes on the
multithreading journey as these changes offer better framework for RFD. See the
mq-async-export branch for current development of one of these prerequisite
changes for multithreading (decoupling export from
Hello!
This may be a bug in protocol selection as we don't care much about builds with
a subset of protocols. Will check it.
Maria
On May 22, 2020 4:55:23 PM GMT+02:00, Nabeel Omer wrote:
>Hi Everyone,
>
>I am trying to build Bird2 with `./configure --with-protocols=bgp` and
>it prints the
You are also missing import all in the first kernel protocol, if I see
correctly.
Maria
On May 5, 2020 1:04:37 AM GMT+02:00, Ondrej Zajicek
wrote:
>On Mon, May 04, 2020 at 11:34:38AM -0700, David Haupt wrote:
>> Hi
>>
>> First time user of bird here so please be gentle. :)
>> I am trying to
Hello!
xmalloc is guaranteed to return non-NULL. If it were to return NULL, BIRD would
die instead. That's why it's xmalloc and not malloc.
Maria
On April 27, 2020 5:26:58 AM GMT+02:00, liupe...@zju.edu.cn wrote:
>Hi,
>
>In lib/string.h line 38,
>
>static inline char *
>xstrdup(const char *c)
function *filter_common_ipv4_out( string remote_peer_name ) {
>> if remote_peer_name = 'peer-neo' then .;
>> }
>
>
>It is stupid because i define the peer-name twice.
>But for now it will do the job relatively easily.
>
>I have over 40 peers, so I have to write the functi
Hello!
Currently not supported. Is simple exact match enough for your purposes? This
may be simply added to the filter language.
Maria
On April 8, 2020 6:47:49 PM GMT+02:00, Michael Rack wrote:
>Hi Guys,
>
>if i am in the EXPORT-Routine, how can i access the "protocol-name" of
>that
>instance
Hello!
Use this list to send a patch set as it comes from git format-patch. You can
also use git send-email to send it.
Maria
On April 6, 2020 12:22:22 PM GMT+02:00, Mattia Milani
wrote:
>Hello to everyone, in the past months I have implemented the MRAI timer
>
>in the BGP protocol and I
as this fix is completely independent on any
recent development.
Maria
On April 5, 2020 1:45:06 AM GMT+02:00, Ross Tajvar wrote:
>Thank you Maria! When will that update reach the Debian source repo, so
>I
>can build the package?
>
>On Sat, Apr 4, 2020 at 7:27 PM Maria Matějka
>
Hello!
Hopefully fixed in current master, commit 2928c5bcc7
The affected test was broken itself, yet it accidentally somehow worked on
little endian.
Maria
On April 4, 2020 1:42:58 AM GMT+02:00, "Maria Matějka"
wrote:
>Hello!
>
>I've just replicated the bug via the proot
Hello!
I've just replicated the bug via the proot approach. Thank you a lot for your
help!
Maria
On April 2, 2020 1:47:32 PM GMT+02:00, Clemens Schrimpe
wrote:
>Hallo all -
>
>I built BIRD (1.x and 2.x) for the EdgeRouter platforms(!) myself for
>many years now and I still do. At first I
The show route command shows only reduced info. Use
show route all
to get full information.
Maria
On April 3, 2020 11:47:04 AM GMT+02:00, Fabiano D'Agostino
wrote:
>Good morning,
>how can I see the as path of a route?
>I think the command is 'show route' and I get:
>192.168.5.0/24 unicast
Hello!
We've not been able to build ourselves on MIPS yet, we went into some strange
problems last time (don't remember exactly). Were you so kind please and could
you please help us setting up Debian for MIPS in QEMU if I fail to manage it
once more?
The main issue was, what hardware to choose
Hello!
> I wonder if there's a good, full example about ROA with static protocol in
> BIRD 2. The following snippet is only for BIRD 1.x, right?
>
> --- %< ---
> roa table roa {
> roa 1.0.0.0/24 max 24 as 13335;
> # […]
> }
> --- %< ---
Yes.
> Based on various snippets and half examples on
Hello!
Well, RFC 5575 doesn't explicitly say that the flowspec rule must contain the
destination chunk, anyway it specifies that these rules should be understood as
additional information for unicast BGP prefixes.
Therefore we assume that the dst is de facto mandatory, despite de iure it is
Hello!
Fixed in master.
Maria
On 2/2/20 5:38 PM, Martin Weinelt wrote:
> Hi,
>
> i was debugging this on and off for the last 36 hours and found the error
> message quite misleading.
>
> [root@ganymede:/etc/bird]# cat bird2.conf
> router id 172.23.42.3;
> protocol device {
> scan time 10;
>
>>> diff --git a/nest/route.h b/nest/route.h
>>> index d2a07f09..b927db5f 100644
>>> --- a/nest/route.h
>>> +++ b/nest/route.h
>>> @@ -458,7 +458,7 @@ typedef struct rta {
>>>protocol-specific metric is
>>> availabe */
>>>
>>>
>>> -const char *
Yes. This version should work.
BTW, I suggest using custom attributes instead of communities for all the route
tagging. They work the same way as other attributes and all protocols (except
for Pipe) ignore them.
Maria
On January 27, 2020 8:52:38 PM GMT+01:00, Robert Blayzor
wrote:
>After
Hello!
You won't advertise anything else than basic IPv4 routes with capabilities off
as the flowspec SAFI needs to be advertised and negotiated by a BGP Capability
Advertisement.
Maria
On 1/21/20 2:05 PM, po...@davidepucci.it wrote:
> Hi.
> I'm setting up Bird (2.0.7) as a BGP FlowSpec
rotocol bgp ege2a_50_4 {
> local as 65533;
> neighbor 1.1.1.1 as 12345;
> source address 206.x.y.z;
>
> description "EDGE2a.50.MIA";
> multihop;
>
> ipv4 {
> table as65533;
> import all
> however, attempts to build it without /--disable-libssh/ result in a linking
> error:
Oops, sorry, I missed one include. Here is the fixed patch, now it compiles
both with and without libSSH.
Maria
diff --git a/configure.ac b/configure.ac
index 40f021a1..da8546a6 100644
--- a/configure.ac
Yes.
To be more specific, e. g. parallel filter processing is already in
development, it even runs, yet it is somehow efficient only for long running
filters. Import/export all is more than 100 times slower for now.
It may happen that some multithreading will be released maybe during 2020.
Aargh, I messed up the reply.
Please send the 1.6 config and 2.0 config and what is wrong. We can't figure
out the cause from this.
Maria
On January 10, 2020 8:58:11 AM GMT+01:00, "Maria Matějka"
wrote:
>Hello!
>
>>I am trying to migrate the config from 1.6 to 2.
Hello!
>I am trying to migrate the config from 1.6 to 2.0 on CentOS 8. I tried
>to follow
>https://gitlab.labs.nic.cz/labs/bird/wikis/transition-notes-to-bird-2
>but my test config didn't work. Is there another guide or wiki I can
>follow to be able to migrate to version 2 ?
What didn't
Hello!
Have you tried to use the workaround mentioned here?
https://bird.network.cz/pipermail/bird-users/2015-August/009882.html
Maria
On December 25, 2019 9:59:30 AM GMT+01:00, "穴澤 正恒" wrote:
>Dear all,
>
>My name is Masatsune Anazawa, an affiliate of JPIX.
>We are verifying the root server
Hello!
How much time does it take to list the kernel table?
time ip r > /dev/null
How many routes do you have in bird table?
show route count
And what export filter do you have for the kernel protocol in bird?
Thanks
Maria
On December 21, 2019 1:07:25 PM GMT+01:00, Nico Schottelius
wrote:
ist;
>ipv4 {
>#import filter {
> # if net ~ [0.0.0.0/0, 192.168.255.0/24] then reject;
># accept;
>#};
># export all;
>import all;
>};
>}
>
>thx
>
>
>
Hello!
It may be GCC; I have just compiled BIRD on GCC 4.7.2 on Debian Wheezy.
If you have an older GCC, it may happen that it doesn't support flattened
struct initializers (which produces this kind of compile errors).
I remember seeing some similar compile error on older GCC's some time ago
Hello!
This happens when bgp_med is initially undefined. Use something like
if defined(bgp_med) then bgp_med = 100; else bgp_med = bgp_med + 100;
Maria
On 11/12/19 11:01 PM, Benedikt Neuffer wrote:
> Hello,
>
> I am trying to increase the MED attribute in an export filter:
> bgp_med =
Hello, this is due to the RPKI table being empty on startup. As a workaround, I
suggest having bgp sessions with delayed start.
The problem is also that bird doesn't reevaluate affected routes after ROA has
changed. This is going to be fixed in near future, yet now the best thing to do
is
0 0 0
> 16275
> Import withdraws:4 0--- 0
> 4
> Export updates: 0 0 0---
> 0
> Export withdraws:0---------
> 0
>
> So h
Hello!
On October 22, 2019 5:51:56 AM GMT+02:00, Darren O'Connor
wrote:
>I was reading the documentation for bird2 when I came across this:
>You can validate routes (RFC 6483) using function roa_check() in filter
>and
>set it as import filter at the BGP protocol. BIRD should re-validate
>all
On September 29, 2019 2:13:00 PM GMT+02:00, Christoph
wrote:
>> This means there are events that take 6 s, possibly postponing other
>> events. May cause problems if you would run some protocols configured
>> with tight timers (e.g. OSPF set to 5 s dead interval).
>>
>> This is probably
Hello!
I would suggest reimplementing what is in sysdep/linux, maybe partly, maybe
completely. The netlink api is implemented in sysdep/linux/netlink.c, the
internal api between BIRD's Unix common code and netlink is not documented; I
would suggest looking for non-static functions in netlink.c
Will check it, thanks for your report.
Maria
On August 24, 2019 12:22:10 AM GMT+02:00, "Дружок 1" wrote:
>Hi!
>..
>/var/log/bird.log
>2019-08-22 15:38:38.604 filters, line 1023: Stack underflow
>2019-08-22 18:02:27.157 filters, line 511: Stack underflow
>2019-08-22 18:02:30.616 filters, line
>It is not a memory you should be concerned about. The main issue of
>bird with huge BGP tables is CPU – bird uses only one core.
Which is what is going to be fixed in a short time (about a month to first test
release, I hope), at least if you don't use Kernel protocol.
Merge to master is
Hello!
This feature seems to be feasible after converting route update path to the
thread safe version. Current optimistic estimation of this feature is Q3 2020,
then the flap dampening should be simple to add.
Maria
On July 23, 2019 11:26:28 PM GMT+02:00, Matthias Waehlisch
wrote:
>Hi,
>
>
Hello!
Have you compiled BIRD yourself, or are you using a package?
This seems to be an "unknown protocol" problem, when RPKI is silently omitted
from your binary due to missing dependencies.
Maria
On July 12, 2019 11:26:51 AM GMT+02:00, Irene Lalioti
wrote:
>Hello dear Maria and bird
Hello!
Please check your filedescriptor limit by
ulimit -n
and check how many files are open by
lsof -p
BIRD has no such limit, it is capable of handling any reasonable (fitting in
memory) number of file descriptors.
If you wanted any other help, feel free to ask more. BIRD should have no
Fine!
Anyway, I can offer you this: Most of the time spent is the configuration of
OVS as I'm not familiar with it. If you could provide me with a script that I
could just run with no configured OVS before and the bug manifests… whooa, it
would really help me.
Maria
On May 7, 2019 6:53:42 AM
Hello!
I don't know about any issue of this kind, anyway we don't test the control
socket to be open for a long time. Are you able to reproduce it in some way?
Thanks
Maria
On April 26, 2019 11:19:50 AM GMT+02:00, Kenth Eriksson
wrote:
>Hi!
>
>We are remote controlling a bird daemon using a
Scheck
wrote:
>Hello Maria,
>
>On Fri, 26 Apr 2019, Maria Matějka wrote:
>> We don't know about any, sorry. It is something we'd like to create
>and maintain. If you are into helping us with this issue, you're very
>welcome.
>
>is there currently any test suite or are t
Hello!
We don't know about any, sorry. It is something we'd like to create and
maintain. If you are into helping us with this issue, you're very welcome.
Maria
On April 26, 2019 12:35:00 AM GMT+02:00, Robert Scheck
wrote:
>Hello all,
>
>is there something like a full (regression) test suite
Or, probably what would be even better – could you please help me running QEMU
with this specific guest architecture? This is definitely an architecture which
we would like to include in our test suite.
Thanks
Maria
On April 25, 2019 11:36:11 PM GMT+02:00, "Maria Matějka"
wro
Hello!
ulimit -c unlimited
bird ...
(after it crashes)
gdb bird core
>bird crashes a few seconds after startup on my Hardkernel ODROID HC-2
>running Ubuntu. Kernel version is 4.14.111-158.
>
>
>How can I debug this?
>
>
>I tried:
>
>
>$ bird -c bird.conf -d -D debug.log
>
On April 15, 2019 3:03:38 PM GMT+02:00, Brian Topping
wrote:
>> On Apr 14, 2019, at 11:07 PM, Maria Matějka
>wrote:
>>
>> And couldn't you just share what IDE you are trying to setup with
>BIRD and what exact problems you have? It should take me less time
>In many cases, it’s sufficient to do a build and then load the
>directory of built sources. Generated source out of something like
>bison or flex would then be parsed as normal. But for some reason, it’s
>not doing that with the BIRD tree. So this inquiry was really about
>whether I would be
>> I think any IDE with C support you like should work, maybe except for
>> Visual Studio which has no real C compiler anyway so it won't build
>> there at all.
>
>How do you figure? Visual Studio has an ANSI C compiler.
Oh, sorry, I always thought that the compiler was buggy even in ANSI
On April 12, 2019 8:52:47 PM GMT+02:00, Brian Topping
wrote:
>Developer team:
>
>I’d like to get better with the source. Due to being an inferior
>specimen and/or too many other projects, I need help from tools like
>CLion to make sense of source trees within weeks instead of years.
>
>Are
Hi!
Did you try using BGP path mask? Like
if (bgp.path ~ [= ? bogon bogon * =]) …
Writing this from my phone. Please consult documentation or filter/test.conf
file in source tree for exact syntax and examples.
Maria
On March 26, 2019 10:34:39 AM GMT+01:00, Alexander Zubkov
wrote:
>Hi,
>
Hello,
if it occurs also on 1.6.4, could you please add also "birdcl show route count"
output and some logs with debug all switched on?
How does it get leaked, do you reconfigure many times? Does the memory
consumption increase gradually or by larger amounts?
Thanks
Maria
Quidquid latine
Try
# ip link add lo1 type dummy
# ip a add 10.10.10.10/24 dev lo1
The ifconfig tool is deprecated.
M.
Quidquid latine dictum sit, altum videtur.
-Original Message-
From: RICCARDO RUSSO
To: bird-users@network.cz
Sent: Thu, 12 Apr 2018 4:41 PM
Subject:
Hello!
First of all, we don't support this old version. Please upgrade, either to
1.6.4 or to 2.0.2.
Second, the other side seems to send multiprotocol NLRI to you. BIRD in 1.x
versions AFAIK does not support this extension and should not announce it. What
is the remote software?
Writing
The bgp_path.last_nonaggregated attribute should help you. Don't have user
manual right now here, writing from my phone, just trying to recall it from my
memory.
Maria
Quidquid latine dictum sit, altum videtur.
-Original Message-
From: Radu Anghel
To:
Bird's internal sprintf doesn't support floats, at least for now, in 1.6.x
probably forever. There may be some patch that adds it. Look for some branch
with "flowspec", I think I was implementing something like this for printing
BGP flowspec and it is probably not merged. I don't have the git
We may also add iface address access in future but it isn't as easy as it
seems. The iface may have more than one address and we still don't have a
method to iterate over a list in filters. Anyway, the "foreach" construction is
on our todo list... Stay tuned!
M.
Quidquid latine dictum sit,
Hi,
there definitely is some segfault fix but there may be more bugs. If you could
send at least your config file, it would be helpful. Anyway, the crash file
(corefile) will help much more.
Thanks
Maria
Quidquid latine dictum sit, altum videtur.
-Original Message-
From: "David S."
84 matches
Mail list logo