[e-nsp] multicast packets reordering

Fabian fabian.extr at gmail.com
Thu Sep 25 05:13:49 EDT 2008


Is it out of order or duplication (I don't know how iperf is handling the
duplication)?

I would say that you should look at the PIM timers on the Redback because in
pure L2 the Extreme are just handling some IGMP snooping and otherwise it is
like flooding.

Regards, Fabian.

-----Original Message-----
From: extreme-nsp-bounces at puck.nether.net
[mailto:extreme-nsp-bounces at puck.nether.net] On Behalf Of Marcin Kuczera
Sent: mardi 23 septembre 2008 19:54
Cc: extreme-nsp at puck.nether.net; redback-nsp at puck.nether.net
Subject: Re: [e-nsp] multicast packets reordering

Fabian wrote:
> Hi Marcin,
> 
> It should not (and if using LAG/Multilink with 'i' Extreme switches, all
> Mcast traffic goes through the same port).
> 
> Are you using them with IP Mcast routing (PIM, DVMRP) or simple Layer 2
> VLAN?
> I would suggest to test without the Redback or connecting your TestSystem
> level by level.

ok, little correction.
Signal provider has removed some flapping MAC (router) from his network, 
so now out-of-order packets look like that:
Sep 22 15:29:15 10.200.180.4 iperf: [ 3] 16600.0-16610.0 sec 1 datagrams
received out-of-order
Sep 22 15:38:05 10.200.180.4 iperf: [ 3] 17130.0-17140.0 sec 1 datagrams
received out-of-order
Sep 22 15:47:05 10.200.180.4 iperf: [ 3] 17670.0-17680.0 sec 1 datagrams
received out-of-order
Sep 22 15:55:55 10.200.180.4 iperf: [ 3] 18200.0-18210.0 sec 1 datagrams
received out-of-order
Sep 22 16:04:45 10.200.180.4 iperf: [ 3] 18730.0-18740.0 sec 1 datagrams
received out-of-order
Sep 22 16:13:45 10.200.180.4 iperf: [ 3] 19270.0-19280.0 sec 1 datagrams
received out-of-order
Sep 22 16:22:35 10.200.180.4 iperf: [ 3] 19800.0-19810.0 sec 1 datagrams
received out-of-order
Sep 22 16:31:25 10.200.180.4 iperf: [ 3] 20330.0-20340.0 sec 1 datagrams
received out-of-order
Sep 22 16:40:15 10.200.180.4 iperf: [ 3] 20860.0-20870.0 sec 1 datagrams
received out-of-order


exaclty every 530seconds !
connectivity is like that:
source-->AlpineA-->RedBack(PIM)-->AlpineA-->AlpineB-->Summit48-->TestSystem.

soft in Alpine 7.6.4.4
soft in RedBack 6.1.1.4


Any Idea what could that be ?

Regards,
Marcin







> 
> I'am testing IP TV using extreme core based on quite old, but
> still alive Alpine3808,3804 and Summit5i switches.
> 
> At the moment, the test device generating multicast reports
> is connected in this way:
> 
>
source-->AlpineA-->RedBack(PIM)-->AlpineA-->AlpineB-->Summit48-->TestSystem.
> 
> TestSystem reports:
> Sep 18 17:41:00 10.200.180.4 iperf: [ 3] 11310.0-11320.0 sec 2 datagrams
> received out-of-order
> Sep 18 17:41:30 10.200.180.4 iperf: [ 3] 11340.0-11350.0 sec 1 datagrams
> received out-of-order
> Sep 18 17:42:10 10.200.180.4 iperf: [ 3] 11380.0-11390.0 sec 2 datagrams
> received out-of-order
> Sep 18 17:42:40 10.200.180.4 iperf: [ 3] 11410.0-11420.0 sec 1 datagrams
> received out-of-order
> Sep 18 17:43:00 10.200.180.4 iperf: [ 3] 11430.0-11440.0 sec 1 datagrams
> received out-of-order
> Sep 18 17:44:20 10.200.180.4 iperf: [ 3] 11510.0-11520.0 sec 1 datagrams
> received out-of-order
> Sep 18 17:45:40 10.200.180.4 iperf: [ 3] 11590.0-11600.0 sec 2 datagrams
> received out-of-order
> Sep 18 17:45:50 10.200.180.4 iperf: [ 3] 11600.0-11610.0 sec 1 datagrams
> received out-of-order
> Sep 18 17:46:00 10.200.180.4 iperf: [ 3] 11610.0-11620.0 sec 3 datagrams
> received out-of-order
> Sep 18 17:46:10 10.200.180.4 iperf: [ 3] 11620.0-11630.0 sec 1 datagrams
> received out-of-order
> Sep 18 17:46:30 10.200.180.4 iperf: [ 3] 11640.0-11650.0 sec 1 datagrams
> received out-of-order
> 
> looks nasty...
> 
> soft in Alpine 7.6.4.4
> soft in RedBack 6.1.1.4
> 
> No multilinks...
> 
> Is there any possibility that Alpine does the reordering of packets ??
> If yes, is there any solution to that problem ?
> 
> Reordering is common problem from the source side, but... I have 10x 
> more reorders that others..
> 
> Regards,
> Marcin
> _______________________________________________
> extreme-nsp mailing list
> extreme-nsp at puck.nether.net
> https://puck.nether.net/mailman/listinfo/extreme-nsp
> 
> 

_______________________________________________
extreme-nsp mailing list
extreme-nsp at puck.nether.net
https://puck.nether.net/mailman/listinfo/extreme-nsp



More information about the extreme-nsp mailing list