Hi Simon,
thanks for the quick reply.
Simon Wunderlich sw@simonwunderlich.de schrieb am 08.02.2016 13:29:55:
Von: Simon Wunderlich sw@simonwunderlich.de An: b.a.t.m.a.n@lists.open-mesh.org Kopie: Andreas Pape APape@phoenixcontact.com Datum: 09.02.2016 07:20 Betreff: Re: [B.A.T.M.A.N.] Looping unicast packets when using BLA
Hi Andreas,
On Monday 08 February 2016 12:35:35 Andreas Pape wrote:
Hello
I have a problem in my mesh setup which is quite similiar to Bug#216
of
the bug tracker. I'm using batman-adv 2014.4.0 in a BLA setup consisting of 3 Mesh
Nodes
(A, B, C) connected to the same backone network via a common switch
and a
mesh node D connected to an end device E. I ping that single mesh node
D
and the connected end device E from a PC which is connected to the
same
switch as the three Nodes A to C. BLA is compiled and enabled.
First of all, did you test v2016.0? v2014.4.0 is pretty old, the bug was
created and closed in 2015 after all ...
I just restarted my last year's work to test batman-adv and was a little bit lazy to update to the latests version as my devices use a fairly old kernel version 2.6.32. And the update to 2014.4.0 early last year only worked with Marek's help (issue in the compat code).
But before making further assumptions, I'll start with the update first. In the meantime I am pretty sure, that the problem does not come from the bla code as such. I changed the code in batadv_bla_rx in the repsective part as follows:
ether_addr_copy(search_claim.addr, ethhdr->h_source); search_claim.vid = vid; claim = batadv_claim_hash_find(bat_priv, &search_claim);
if (!claim) { /* possible optimization: race for a claim */ /* No claim exists yet, claim it for us! */
if (!batadv_is_my_client(bat_priv, ethhdr->h_source, vid)) { batadv_handle_claim(bat_priv, primary_if, primary_if->net_dev->dev_addr, ethhdr->h_source, vid); goto allow; } else { printk("not claimed: %pM \n", ethhdr->h_source); goto handled; } }
I did this yesterday in a "quick-and-dirty" way and restarted my pingtest, which ran until this morning without looping packets. But I did not notice until now that I did not only prevent the claiming of MAC addresses from the own backbone but I also dropped the packets causing the claim to be triggered! That tells me that the original code in batadv_bla_rx is most likely OK and that my problem comes from somewhere else (e.g. ping request from PC to device E enters gateway A and is forwarded to gateway B via the mesh. But gateway B does not forward it to mesh node D but sends the packet via the linux bridge and my eth0 interface to the backbone network).
But before digging deeper into this, I'll make a try with 2016.0 and see if the problem is solved there.
From time to time I see looping unicast packets in my backbone
network.
This unicast looping starts directly after one of the nodes A to C
claimed
the mac address of my PC. The looping telegram is then the ping
request
sent by the PC. I have a wireshark recording made in my backbone via
port
mirroring of one of the switch ports where a mesh node is connected to which shows this behaviour.
Is it really the ping packet looping? If yes, which nodes are part of
the
loop? Normally we only see broadcast packets looping. In #216 it was
also
broadcast packets where we have seen duplicates, and this was more a
locking
problem leading to creation of the same packets again and again.
I am not sure if I understood bla correctly but isn't it nonsense that
a
bla backbone gateway claims MAC addresses from its own backbone (i.e.
the
one it is directly connected to via its ethernet port)?
Yes, that appears to be nonsense indeed. Do you happen to have DAT
enabled?
There were also some problems with that which are fixed by now.
DAT is enabled. But my problem starts with a gratuitous arp containing a claim and not a multiplication of normal arp requests or repsonses.
A simple change in batadv_bla_rx seems to solve this problem: add an additional check before claiming a new mac address: if this address is already known from the tt local table (via command
batadv_is_my_client)
don't claim it.
This seems to solve my problem as far as I have tested so far. Any thoughts about that?
This will prevent roaming from on of your nodes connected to the
backbone (A-
C) to the mesh-only node D.
I would like to suggest to upgrade and test again, and try disabling DAT
if
the problem is still present (you should still report it if DAT makes a difference in that case). If you still see a problem then, we probably
have
something unsolved, and then I'd like to understand which nodes are part
of
the loop.
Thank you! Simon[Anhang "signature.asc" gelöscht von Andreas Pape/Phoenix
Contact]
Thanks and regards, Andreas
.................................................................. PHOENIX CONTACT ELECTRONICS GmbH
Sitz der Gesellschaft / registered office of the company: 31812 Bad Pyrmont USt-Id-Nr.: DE811742156 Amtsgericht Hannover HRB 100528 / district court Hannover HRB 100528 Geschäftsführer / Executive Board: Roland Bent, Dr. Martin Heubeck ___________________________________________________________________ Diese E-Mail enthält vertrauliche und/oder rechtlich geschützte Informationen. Wenn Sie nicht der richtige Adressat sind oder diese E-Mail irrtümlich erhalten haben, informieren Sie bitte sofort den Absender und vernichten Sie diese Mail. Das unerlaubte Kopieren, jegliche anderweitige Verwendung sowie die unbefugte Weitergabe dieser Mail ist nicht gestattet. ---------------------------------------------------------------------------------------------------- This e-mail may contain confidential and/or privileged information. If you are not the intended recipient (or have received this e-mail in error) please notify the sender immediately and destroy this e-mail. Any unauthorized copying, disclosure, distribution or other use of the material or parts thereof is strictly forbidden. ___________________________________________________________________
Hi,
I want to give a short feedback concerning my attempt to use batman-adv-2016.0 instead of to older version I used first.
Unfortunately, using batman-adv-2016.0 does not solve my problem. I can still see claim frames sent by the mesh gateways into the common backbone network for MAC addresses out of the common backbone network itself. I enabled again both bla and dat support. Furthermore I have also still problems with DAT, because there are multiple ARP replies visible in the backbone network coming out of the mesh.
That reminds me that I have forgotten to mention in my earlier mail, that I did not test with the original batman-adv-2014.4.0 first but with a version having the patch applied I sent to the mailing list in March last year concerning possible fixes for dat in bla setups. If I remember it correctly I think there were two main issues in batman-adv-2014.4.0 when using dat in combination with bla: 1.Broadcast ARP requests from the backbone network are handled by each gateway, leading to multiple dat adress resoultions in parallel. 2. As dat uses tunneling of broadcasts in special batman-adv unicast frames, the current bla code does not seem to prevent these broadcasts from reaching the backone network as it is done for normal broadcast coming from the mesh and heading for the backbone. Both effects together lead to a multiplication of arp requests and replies. My patch of last year tried to address this.
Good news is that disabling dat in batman-adv-2016.0 seems to solve my observed issues (in strange ways even the observed erroneous claim frames in the backbone network....). But I think dat is a clever feature to reduce broadcast load in the mesh network. Wouldn't it be useful to dig a little bit deeper into the combined use of dat and bla? I would volunteer for testing and providing ideas for improving the behaviour.
Or do you think that I have an issue with my old 2.6.32 kernel?
Best regards, Andreas
"B.A.T.M.A.N" b.a.t.m.a.n-bounces@lists.open-mesh.org schrieb am 09.02.2016 08:01:27:
Von: Andreas Pape APape@phoenixcontact.com An: Simon Wunderlich sw@simonwunderlich.de Kopie: b.a.t.m.a.n@lists.open-mesh.org Datum: 09.02.2016 08:01 Betreff: [B.A.T.M.A.N.] Antwort: Re: Looping unicast packets when using
BLA
Gesendet von: "B.A.T.M.A.N" b.a.t.m.a.n-bounces@lists.open-mesh.org
Hi Simon,
thanks for the quick reply.
Simon Wunderlich sw@simonwunderlich.de schrieb am 08.02.2016 13:29:55:
Von: Simon Wunderlich sw@simonwunderlich.de An: b.a.t.m.a.n@lists.open-mesh.org Kopie: Andreas Pape APape@phoenixcontact.com Datum: 09.02.2016 07:20 Betreff: Re: [B.A.T.M.A.N.] Looping unicast packets when using BLA
Hi Andreas,
On Monday 08 February 2016 12:35:35 Andreas Pape wrote:
Hello
I have a problem in my mesh setup which is quite similiar to Bug#216
of
the bug tracker. I'm using batman-adv 2014.4.0 in a BLA setup consisting of 3 Mesh
Nodes
(A, B, C) connected to the same backone network via a common switch
and a
mesh node D connected to an end device E. I ping that single mesh
node
D
and the connected end device E from a PC which is connected to the
same
switch as the three Nodes A to C. BLA is compiled and enabled.
First of all, did you test v2016.0? v2014.4.0 is pretty old, the bug
was
created and closed in 2015 after all ...
I just restarted my last year's work to test batman-adv and was a little bit lazy to update to the latests version as my devices use a fairly old kernel version 2.6.32. And the update to 2014.4.0 early last year only worked with Marek's help (issue in the compat code).
But before making further assumptions, I'll start with the update first. In the meantime I am pretty sure, that the problem does not come from
the
bla code as such. I changed the code in batadv_bla_rx in the repsective part as follows:
ether_addr_copy(search_claim.addr, ethhdr->h_source); search_claim.vid = vid; claim = batadv_claim_hash_find(bat_priv, &search_claim); if (!claim) { /* possible optimization: race for a claim */ /* No claim exists yet, claim it for us! */ if (!batadv_is_my_client(bat_priv, ethhdr->h_source,
vid))
{ batadv_handle_claim(bat_priv, primary_if, primary_if->net_dev->dev_addr, ethhdr->h_source, vid); goto allow; } else { printk("not claimed: %pM \n", ethhdr->h_source); goto handled; } }
I did this yesterday in a "quick-and-dirty" way and restarted my
pingtest,
which ran until this morning without looping packets. But I did not
notice
until now that I did not only prevent the claiming of MAC addresses from the own backbone but I also dropped the packets causing the claim to be triggered! That tells me that the original code in batadv_bla_rx is most likely OK and that my problem comes from somewhere else (e.g. ping
request
from PC to device E enters gateway A and is forwarded to gateway B via
the
mesh. But gateway B does not forward it to mesh node D but sends the packet via the linux bridge and my eth0 interface to the backbone network).
But before digging deeper into this, I'll make a try with 2016.0 and see if the problem is solved there.
From time to time I see looping unicast packets in my backbone
network.
This unicast looping starts directly after one of the nodes A to C
claimed
the mac address of my PC. The looping telegram is then the ping
request
sent by the PC. I have a wireshark recording made in my backbone via
port
mirroring of one of the switch ports where a mesh node is connected
to
which shows this behaviour.
Is it really the ping packet looping? If yes, which nodes are part of
the
loop? Normally we only see broadcast packets looping. In #216 it was
also
broadcast packets where we have seen duplicates, and this was more a
locking
problem leading to creation of the same packets again and again.
I am not sure if I understood bla correctly but isn't it nonsense
that
a
bla backbone gateway claims MAC addresses from its own backbone
(i.e.
the
one it is directly connected to via its ethernet port)?
Yes, that appears to be nonsense indeed. Do you happen to have DAT
enabled?
There were also some problems with that which are fixed by now.
DAT is enabled. But my problem starts with a gratuitous arp containing a claim and not a multiplication of normal arp requests or repsonses.
A simple change in batadv_bla_rx seems to solve this problem: add an additional check before claiming a new mac address: if this address
is
already known from the tt local table (via command
batadv_is_my_client)
don't claim it.
This seems to solve my problem as far as I have tested so far. Any thoughts about that?
This will prevent roaming from on of your nodes connected to the
backbone (A-
C) to the mesh-only node D.
I would like to suggest to upgrade and test again, and try disabling
DAT
if
the problem is still present (you should still report it if DAT makes
a
difference in that case). If you still see a problem then, we probably
have
something unsolved, and then I'd like to understand which nodes are
part
of
the loop.
Thank you! Simon[Anhang "signature.asc" gelöscht von Andreas Pape/Phoenix
Contact]
Thanks and regards, Andreas
.................................................................. PHOENIX CONTACT ELECTRONICS GmbH
Sitz der Gesellschaft / registered office of the company: 31812 Bad
Pyrmont
USt-Id-Nr.: DE811742156 Amtsgericht Hannover HRB 100528 / district court Hannover HRB 100528 Geschäftsführer / Executive Board: Roland Bent, Dr. Martin Heubeck ___________________________________________________________________ Diese E-Mail enthält vertrauliche und/oder rechtlich geschützte Informationen. Wenn Sie nicht der richtige Adressat sind oder diese E-Mail irrtümlich erhalten haben, informieren Sie bitte sofort den Absender und vernichten Sie diese Mail. Das unerlaubte Kopieren, jegliche anderweitige Verwendung sowie die unbefugte Weitergabe dieser Mail ist nicht gestattet.
----------------------------------------------------------------------------------------------------
This e-mail may contain confidential and/or privileged information. If you are not the intended recipient (or have received this e-mail in error) please notify the sender immediately and destroy this e- mail. Any unauthorized copying, disclosure, distribution or other use of the material or parts thereof is strictly forbidden. ___________________________________________________________________
.................................................................. PHOENIX CONTACT ELECTRONICS GmbH
Sitz der Gesellschaft / registered office of the company: 31812 Bad Pyrmont USt-Id-Nr.: DE811742156 Amtsgericht Hannover HRB 100528 / district court Hannover HRB 100528 Geschäftsführer / Executive Board: Roland Bent, Dr. Martin Heubeck ___________________________________________________________________ Diese E-Mail enthält vertrauliche und/oder rechtlich geschützte Informationen. Wenn Sie nicht der richtige Adressat sind oder diese E-Mail irrtümlich erhalten haben, informieren Sie bitte sofort den Absender und vernichten Sie diese Mail. Das unerlaubte Kopieren, jegliche anderweitige Verwendung sowie die unbefugte Weitergabe dieser Mail ist nicht gestattet. ---------------------------------------------------------------------------------------------------- This e-mail may contain confidential and/or privileged information. If you are not the intended recipient (or have received this e-mail in error) please notify the sender immediately and destroy this e-mail. Any unauthorized copying, disclosure, distribution or other use of the material or parts thereof is strictly forbidden. ___________________________________________________________________
Hi Andreas,
On Thursday 11 February 2016 10:19:07 Andreas Pape wrote:
Hi,
I want to give a short feedback concerning my attempt to use batman-adv-2016.0 instead of to older version I used first.
Unfortunately, using batman-adv-2016.0 does not solve my problem. I can still see claim frames sent by the mesh gateways into the common backbone network for MAC addresses out of the common backbone network itself. I enabled again both bla and dat support. Furthermore I have also still problems with DAT, because there are multiple ARP replies visible in the backbone network coming out of the mesh.
That reminds me that I have forgotten to mention in my earlier mail, that I did not test with the original batman-adv-2014.4.0 first but with a version having the patch applied I sent to the mailing list in March last year concerning possible fixes for dat in bla setups. If I remember it correctly I think there were two main issues in batman-adv-2014.4.0 when using dat in combination with bla: 1.Broadcast ARP requests from the backbone network are handled by each gateway, leading to multiple dat adress resoultions in parallel.
That shouldn't be a problem on its own.
- As dat uses tunneling of broadcasts in special batman-adv unicast
frames, the current bla code does not seem to prevent these broadcasts from reaching the backone network as it is done for normal broadcast coming from the mesh and heading for the backbone. Both effects together lead to a multiplication of arp requests and replies. My patch of last year tried to address this.
Hm, I see. I just checked the code and it seems we fixed this issue for speedy join in the mean time (affecting TT), but for bla the problem is still present.
Wouldn't it be sufficient to add something like a check for backbones ( batadv_bla_is_backbone_gw) into batadv_recv_unicast_packet() and drop packets if they came from the same backbone?
I have found your patch from last year. Would you like to rebase/split your patch to address the remaining issues? That would help us a lot. Please also put a proper patch message. I promise to be more responsive this time. :)
Good news is that disabling dat in batman-adv-2016.0 seems to solve my observed issues (in strange ways even the observed erroneous claim frames in the backbone network....). But I think dat is a clever feature to reduce broadcast load in the mesh network. Wouldn't it be useful to dig a little bit deeper into the combined use of dat and bla? I would volunteer for testing and providing ideas for improving the behaviour.
If you could help, that would be great! Your patch from last year already is a good start, so I'm sure you are capable of working on that. I would be happy to help with that, too.
Or do you think that I have an issue with my old 2.6.32 kernel?
I don't think so. To me it looks like the issue could still be present in current versions ...
Cheers, Simon
Hi Simon,
Simon Wunderlich sw@simonwunderlich.de schrieb am 11.02.2016 12:08:25:
Von: Simon Wunderlich sw@simonwunderlich.de An: Andreas Pape APape@phoenixcontact.com Kopie: The list for a Better Approach To Mobile Ad-hoc Networking b.a.t.m.a.n@lists.open-mesh.org, "B.A.T.M.A.N" <b.a.t.m.a.n- bounces@lists.open-mesh.org> Datum: 11.02.2016 12:08 Betreff: Re: Re: [B.A.T.M.A.N.] Antwort: Re: Looping unicast packets when using BLA
Hi Andreas,
On Thursday 11 February 2016 10:19:07 Andreas Pape wrote:
Hi,
I want to give a short feedback concerning my attempt to use batman-adv-2016.0 instead of to older version I used first.
Unfortunately, using batman-adv-2016.0 does not solve my problem. I
can
still see claim frames sent by the mesh gateways into the common
backbone
network for MAC addresses out of the common backbone network itself. I enabled again both bla and dat support. Furthermore I have also still problems with DAT, because there are multiple ARP replies visible in
the
backbone network coming out of the mesh.
That reminds me that I have forgotten to mention in my earlier mail,
that
I did not test with the original batman-adv-2014.4.0 first but with a version having the patch applied I sent to the mailing list in March
last
year concerning possible fixes for dat in bla setups. If I remember it correctly I think there were two main issues in batman-adv-2014.4.0
when
using dat in combination with bla: 1.Broadcast ARP requests from the backbone network are handled by each gateway, leading to multiple dat adress resoultions in parallel.
That shouldn't be a problem on its own.
I think I wasn't precise enough concerning this point. I meant the effect, that a broadcast ARP coming from a common backbone reaches all gateways. If now accidentally several gateways can already answer that request due to dat, then the current code sends an arp reply from each gateway being able to answer. This broadcast does not even reach the mesh if all gateways can answer the request (as far as I have understood the code). Therefore broadcast handling in the mesh layer does not solve this problem.
- As dat uses tunneling of broadcasts in special batman-adv unicast
frames, the current bla code does not seem to prevent these broadcasts from reaching the backone network as it is done for normal broadcast coming from the mesh and heading for the backbone. Both effects together lead to a multiplication of arp requests and replies. My patch of last year tried to address this.
Hm, I see. I just checked the code and it seems we fixed this issue for speedy join in the mean time (affecting TT), but for bla the problem is still present.
Wouldn't it be sufficient to add something like a check for backbones ( batadv_bla_is_backbone_gw) into batadv_recv_unicast_packet() and drop
packets
if they came from the same backbone?
That's a good question. This is something I did not dare to do last year because I cannot foresee possible negative implications. Perhaps someone more experienced with batman-adv routing should answer this. Therefore I focussed on fixing this for the DAT ARP handling only. But doing so as you propose would most likely have the positive side effect that I will get rid of the looping unicast packets in my backbone network, too. But as mentioned in my prior mail I think this looping unicast packets might have another cause. The question is: why does a gateway forward a unicast packet received via the mesh with a destination mac behind an originator somewhere else in the mesh (that originator is not connected to the same backbone) to its own backbone? If this only can happen if the entry in the global tt expires (like a mac adress table expires in a switch and the switch starts broadcasting incoming packets to all ports), then I would think blocking all unicast traffic coming from another backbone gw of the same backbone network is the smartest and easiest solution.
I have found your patch from last year. Would you like to rebase/split
your
patch to address the remaining issues? That would help us a lot. Please
also
put a proper patch message. I promise to be more responsive this time.
:)
I'm trying to split my last year's patch into logical seperate pieces, update them to be compliant to the latest master branch of the batman-adv.git repositor and will mail them for further discussion.
Good news is that disabling dat in batman-adv-2016.0 seems to solve my observed issues (in strange ways even the observed erroneous claim
frames
in the backbone network....). But I think dat is a clever feature to reduce broadcast load in the mesh network. Wouldn't it be useful to
dig a
little bit deeper into the combined use of dat and bla? I would
volunteer
for testing and providing ideas for improving the behaviour.
If you could help, that would be great! Your patch from last year already is a good start, so I'm sure you are capable of working on that. I would be
happy
to help with that, too.
Or do you think that I have an issue with my old 2.6.32 kernel?
I don't think so. To me it looks like the issue could still be present
in
current versions ...
Cheers, Simon[Anhang "signature.asc" gelöscht von Andreas Pape/Phoenix
Contact]
.................................................................. PHOENIX CONTACT ELECTRONICS GmbH
Sitz der Gesellschaft / registered office of the company: 31812 Bad Pyrmont USt-Id-Nr.: DE811742156 Amtsgericht Hannover HRB 100528 / district court Hannover HRB 100528 Geschäftsführer / Executive Board: Roland Bent, Dr. Martin Heubeck ___________________________________________________________________ Diese E-Mail enthält vertrauliche und/oder rechtlich geschützte Informationen. Wenn Sie nicht der richtige Adressat sind oder diese E-Mail irrtümlich erhalten haben, informieren Sie bitte sofort den Absender und vernichten Sie diese Mail. Das unerlaubte Kopieren, jegliche anderweitige Verwendung sowie die unbefugte Weitergabe dieser Mail ist nicht gestattet. ---------------------------------------------------------------------------------------------------- This e-mail may contain confidential and/or privileged information. If you are not the intended recipient (or have received this e-mail in error) please notify the sender immediately and destroy this e-mail. Any unauthorized copying, disclosure, distribution or other use of the material or parts thereof is strictly forbidden. ___________________________________________________________________
Hi Andreas,
On Friday 12 February 2016 11:40:21 Andreas Pape wrote:
[...] using dat in combination with bla: 1.Broadcast ARP requests from the backbone network are handled by each gateway, leading to multiple dat adress resoultions in parallel.
That shouldn't be a problem on its own.
I think I wasn't precise enough concerning this point. I meant the effect, that a broadcast ARP coming from a common backbone reaches all gateways. If now accidentally several gateways can already answer that request due to dat, then the current code sends an arp reply from each gateway being able to answer. This broadcast does not even reach the mesh if all gateways can answer the request (as far as I have understood the code). Therefore broadcast handling in the mesh layer does not solve this problem.
Yes, we may have multiple gateways answering with an ARP reply. But how is this a problem? It is redundant, yes, but its just a unicast sent back. I don't see this a s problem yet ...
- As dat uses tunneling of broadcasts in special batman-adv unicast
frames, the current bla code does not seem to prevent these broadcasts from reaching the backone network as it is done for normal broadcast coming from the mesh and heading for the backbone. Both effects together lead to a multiplication of arp requests and replies. My patch of last year tried to address this.
Hm, I see. I just checked the code and it seems we fixed this issue for speedy join in the mean time (affecting TT), but for bla the problem is still present.
Wouldn't it be sufficient to add something like a check for backbones ( batadv_bla_is_backbone_gw) into batadv_recv_unicast_packet() and drop
packets
if they came from the same backbone?
That's a good question. This is something I did not dare to do last year because I cannot foresee possible negative implications. Perhaps someone more experienced with batman-adv routing should answer this. Therefore I focussed on fixing this for the DAT ARP handling only. But doing so as you propose would most likely have the positive side effect that I will get rid of the looping unicast packets in my backbone network, too. But as mentioned in my prior mail I think this looping unicast packets might have another cause. The question is: why does a gateway forward a unicast packet received via the mesh with a destination mac behind an originator somewhere else in the mesh (that originator is not connected to the same backbone) to its own backbone? If this only can happen if the entry in the global tt expires (like a mac adress table expires in a switch and the switch starts broadcasting incoming packets to all ports), then I would think blocking all unicast traffic coming from another backbone gw of the same backbone network is the smartest and easiest solution.
Yeah, agreed. There shouldn't be any unicast messages to be sent among gateways through the mesh. We probably should not only avoid the receiving (as I suggested only), but also sending DAT requests to other gateways on the same backbone. The check should be similar and simple ... After all, if there is another gateway capable of answering, it will also receive the request and doesn't need it passed from somebody else on the same backbone.
Regarding the TT question/expiration, as a receiver we don't check the destination address and just accept the packet for receiption. But as I said, packets among gateways on the same backbone shouldn't be sent or received.
I have found your patch from last year. Would you like to rebase/split
your
patch to address the remaining issues? That would help us a lot. Please
also
put a proper patch message. I promise to be more responsive this time.
: :)
I'm trying to split my last year's patch into logical seperate pieces, update them to be compliant to the latest master branch of the batman-adv.git repositor and will mail them for further discussion.
Cool, thanks! I'm looking forward to it. :)
Cheers, Simon
Simon Wunderlich sw@simonwunderlich.de schrieb am 12.02.2016 14:04:23:
Von: Simon Wunderlich sw@simonwunderlich.de An: Andreas Pape APape@phoenixcontact.com Kopie: The list for a Better Approach To Mobile Ad-hoc Networking b.a.t.m.a.n@lists.open-mesh.org Datum: 12.02.2016 14:04 Betreff: Re: Antwort: Re: Re: [B.A.T.M.A.N.] Antwort: Re: Looping unicast packets when using BLA
Hi Andreas,
On Friday 12 February 2016 11:40:21 Andreas Pape wrote:
[...] using dat in combination with bla: 1.Broadcast ARP requests from the backbone network are handled by
each
gateway, leading to multiple dat adress resoultions in parallel.
That shouldn't be a problem on its own.
I think I wasn't precise enough concerning this point. I meant the
effect,
that a broadcast ARP coming from a common backbone reaches all gateways. If
now
accidentally several gateways can already answer that request due to dat, then the current code sends an arp reply from each gateway being able to answer. This broadcast
does
not even reach the mesh if all gateways can answer the request (as far as I have understood the code). Therefore broadcast handling in the mesh layer does not solve this problem.
Yes, we may have multiple gateways answering with an ARP reply. But how
is
this a problem? It is redundant, yes, but its just a unicast sent back.
I
don't see this a s problem yet ...
I would like to prevent duplicated packets as much as possible, even if they are unicast packets normally harmlexs for typical PC hardware. But I know of enough small embedded devices (sensors and stuff like that) which don't like that.....
- As dat uses tunneling of broadcasts in special batman-adv
unicast
frames, the current bla code does not seem to prevent these
broadcasts
from reaching the backone network as it is done for normal
broadcast
coming from the mesh and heading for the backbone. Both effects together lead to a multiplication of arp requests and replies. My patch of last year tried to address this.
Hm, I see. I just checked the code and it seems we fixed this issue for speedy join in the mean time (affecting TT), but for bla the problem is
still
present.
Wouldn't it be sufficient to add something like a check for
backbones (
batadv_bla_is_backbone_gw) into batadv_recv_unicast_packet() and
drop
packets
if they came from the same backbone?
That's a good question. This is something I did not dare to do last
year
because I cannot foresee possible negative implications. Perhaps someone more experienced with batman-adv routing should answer this. Therefore I focussed on fixing
this
for the DAT ARP handling only. But doing so as you propose would most likely have the positive side effect that I will get rid of the looping unicast packets in my backbone network, too. But as mentioned in my prior mail I think this looping unicast packets might have another cause. The question is: why does a gateway forward a unicast packet received via the mesh with a destination mac behind an originator somewhere else in the mesh (that originator is not
connected
to the same backbone) to its own backbone? If this only can happen if the entry in the global
tt
expires (like a mac adress table expires in a switch and the switch starts broadcasting incoming packets to all ports), then I would think blocking all unicast traffic coming from another
backbone gw
of the same backbone network is the smartest and easiest solution.
Yeah, agreed. There shouldn't be any unicast messages to be sent among gateways through the mesh. We probably should not only avoid the receiving (as I suggested only), but also sending DAT requests to other gateways on the same backbone. The check should be similar and simple ... After all, if there
is
another gateway capable of answering, it will also receive the request
and
doesn't need it passed from somebody else on the same backbone.
Regarding the TT question/expiration, as a receiver we don't check the destination address and just accept the packet for receiption. But as I
said,
packets among gateways on the same backbone shouldn't be sent or
received.
I have found your patch from last year. Would you like to
rebase/split
your
patch to address the remaining issues? That would help us a lot.
Please
also
put a proper patch message. I promise to be more responsive this
time.
: :)
I'm trying to split my last year's patch into logical seperate pieces, update them to be compliant to the latest master branch of the batman-adv.git repositor and will mail them for further discussion.
Cool, thanks! I'm looking forward to it. :)
I've just sent the patches. They have the state of my "experiments" last year. That means that your latest proposal is not integrated yet. I quickly updated my devices in my test setup and it looks good (no looping arp requests or multiple replies seen so far).
Regards, Andreas
.................................................................. PHOENIX CONTACT ELECTRONICS GmbH
Sitz der Gesellschaft / registered office of the company: 31812 Bad Pyrmont USt-Id-Nr.: DE811742156 Amtsgericht Hannover HRB 100528 / district court Hannover HRB 100528 Geschäftsführer / Executive Board: Roland Bent, Dr. Martin Heubeck ___________________________________________________________________ Diese E-Mail enthält vertrauliche und/oder rechtlich geschützte Informationen. Wenn Sie nicht der richtige Adressat sind oder diese E-Mail irrtümlich erhalten haben, informieren Sie bitte sofort den Absender und vernichten Sie diese Mail. Das unerlaubte Kopieren, jegliche anderweitige Verwendung sowie die unbefugte Weitergabe dieser Mail ist nicht gestattet. ---------------------------------------------------------------------------------------------------- This e-mail may contain confidential and/or privileged information. If you are not the intended recipient (or have received this e-mail in error) please notify the sender immediately and destroy this e-mail. Any unauthorized copying, disclosure, distribution or other use of the material or parts thereof is strictly forbidden. ___________________________________________________________________
Hi Andreas,
On Friday 12 February 2016 15:07:59 Andreas Pape wrote:
Simon Wunderlich sw@simonwunderlich.de schrieb am 12.02.2016 14:04:23:
Von: Simon Wunderlich sw@simonwunderlich.de An: Andreas Pape APape@phoenixcontact.com Kopie: The list for a Better Approach To Mobile Ad-hoc Networking b.a.t.m.a.n@lists.open-mesh.org Datum: 12.02.2016 14:04 Betreff: Re: Antwort: Re: Re: [B.A.T.M.A.N.] Antwort: Re: Looping unicast packets when using BLA
Hi Andreas,
On Friday 12 February 2016 11:40:21 Andreas Pape wrote:
[...] using dat in combination with bla: 1.Broadcast ARP requests from the backbone network are handled by
each
gateway, leading to multiple dat adress resoultions in parallel.
That shouldn't be a problem on its own.
I think I wasn't precise enough concerning this point. I meant the
effect,
that a broadcast ARP coming from a common backbone reaches all gateways. If
now
accidentally several gateways can already answer that request due to dat, then the current code sends an arp reply from each gateway being able to answer. This broadcast
does
not even reach the mesh if all gateways can answer the request (as far as I have understood the code). Therefore broadcast handling in the mesh layer does not solve this problem.
Yes, we may have multiple gateways answering with an ARP reply. But how
is
this a problem? It is redundant, yes, but its just a unicast sent back.
I
don't see this a s problem yet ...
I would like to prevent duplicated packets as much as possible, even if they are unicast packets normally harmlexs for typical PC hardware. But I know of enough small embedded devices (sensors and stuff like that) which don't like that.....
Thats a good point. In general it could be debated whether we prefer redundant replies to no replies at all. But I'd agree to your point, especially since having answers from different devices may confuse a switch because it thinks there is some mac flapping or worse, having answers from different ports.
[...]
I've just sent the patches. They have the state of my "experiments" last year. That means that your latest proposal is not integrated yet. I quickly updated my devices in my test setup and it looks good (no looping arp requests or multiple replies seen so far).
Thanks a lot! I've reviewed them, we still have some formatting work to do so please bear with us with the iterations. Splitting and cleaning them up was definitely a great start, this is a very good contribution. :)
Thanks, Simon
b.a.t.m.a.n@lists.open-mesh.org
-
Andreas Pape -
Simon Wunderlich