Multicast Issue - Out of Sequence Packets

[mamoser]

Hi,

First off I'd like to thank all those who will be helping me through my dilema and provide my current setup:

video server / smart bit (used for testing - outputs multicast streams)
|
cisco 2900 (switch) --> multicast flooded to encapsulator
|
encapsulator --> converted into dvb packets
|
modulator
|
sat dish (upconverting)
|
Satellite (specific transponder)
|
sat dish (downconverting)
|
Satellite Receivers (4 of them - outputs multicast back on another LAN)
||||
cisco 4948 (Layer 3 Switch)
|
client's requesting multicast (Smartbits used again for testing - smartbits sending igmp join request to all multicast addresses)

Description of what i'm trying to accomplish:

I need to send a video/audio feed from one province to another via satellite where client's will be using igmp to request the feeds.

Broken down steps
-smartbit outputs multicast test stream (230.0.0.1, 230.0.0.2, 230.0.0.3, 230.0.0.4)
-stream is flooded onto 2900 switch which the encapsulator is listening
-encapsulator converts ip packets into dvb packets which are then modulated into a specific frequency to my upconverter which simulates my satellite connection.
-sat receivers receive the packets via sat0 interface and outputs the traffic to the cisco 4948.
-client's send igmp join requests to get the video stream but for now i use the smartbit controller to do this.

I have not implemented ip pim nor cgmp for controlling multicast. (when i try i get the same results)

Description of the problem:
I can get the multicast traffic through the satellite connection but my smart bits is detecting a lot of out of sequence packets. I need to find a way to reduce the amount of out of sequence packets but I cannot figure out what's the cause. When I only output one multicast stream through the smart bits, the receiving end has no problems. (1 or 2 out of sequence packets which is normal from what i've been told). The minute I add more streams, I have a serious issue with out of sequence packets.

My co-worker hates cisco and blames the switch. To prove to me that he was right, he took a linksys 24 port switch (igmp passthrough enable) and replaced the 4948 with it. Low and behold, 4 streams with hardly any out of sequence packets. So the issue has been confirmed, it's 4948 switch.

Anyone know what needs to be enable on the 4948 to solve my problem? I'm CCNP certified with a huge background in networking. I've read tones of documentation on multicasting with cisco but nothing regarding the problem i'm facing. Plz I need to get this setup working without using a linksys 24 port switch ASAP.

Hope to hear from someone,

Evan

PS. IF you need more info ask for it or you can e-mail me directly at mamosere@hotmail.com.

 

[Minue]

Hello
Try turning on IGMP snooping or CGMP.
Regards

 

[BuckWeet]

I'm failing to see the issue.. You have 4 seperate multicast streams..

Since they're seperate groups what does it matter if the packets are out of order? As long as they come in sequence for the respective group there is no issue..

I would enable PIM on that vlan (while PIM isn't needed, a subfeature is that will enable IGMP on that vlan). Also verify that you have IGMP snooping enabled so that you are not broadcasting frames out all ports..


BuckWeet

 

[mamoser]

Thanks for the reply, I really appreciate all the help. Now you've mentioned that out of sequence packets should not be an issue but my engineers stats that it is since the traffic is very sensitive to delay. Could you point me to information about out of sequence packets that I could present to my project manager proving that this should not be a concern. I myself doubt the engeneers on this one.

I'm also a little perplexed as to why a linksys L2 switch did not cause this problem.

Here's a little more info:
There will be about 7-15 receivers connected to the 4948 l3 switch. Each receiver will be outputting 70mbps of video/audio multicast broadcasting packets. PIM and CGMP will be used for routing and controlling the multicast traffic.

I've tried enabling PIM and cgmp but it did not solve the problem. I'll look into enabling igmp snooping but I was told not to use it because it will use too much processing on the L3 switch.

When I hard coded both the duplex and speed on both the layer 3 switch and my receivers, the out of sequence packets dropped dramastically.

out of 200 000 packets out of both receivers
400 and 600 are the out of sequence values after hardcoding speed and duplex. I was getting in the hundred thousands before the hard coding.

*linksys had 2 and 4 out of 200 000 packets

I was then recommended the config listed in the next posting by some random guy on the internet but i don't think it works. When i use this config, my smartbits did not see the multicast traffic even though it was sending igmp join messages for the specific mgroup.

I confirmed that the traffic was coming through on the interface g1/1 and g1/5.

GigabitEthernet1/1 is up, line protocol is up (connected)
Hardware is Gigabit Ethernet Port, address is 001e.f71a.5940 (bia 001e.f71a.59
40)
MTU 1500 bytes, BW 100000 Kbit, DLY 10 usec,
reliability 255/255, txload 1/255, rxload 10/255
Encapsulation ARPA, loopback not set
Keepalive set (10 sec)
Full-duplex, 100Mb/s, link type is auto, media type is 10/100/1000-TX
input flow-control is off, output flow-control is off
ARP type: ARPA, ARP Timeout 04:00:00
Last input never, output never, output hang never
Last clearing of "show interface" counters 00:35:50
Input queue: 0/2000/0/0 (size/max/drops/flushes); Total output drops: 0
Queueing strategy: fifo
Output queue: 0/40 (size/max)
5 minute input rate 4302000 bits/sec, 706 packets/sec
5 minute output rate 0 bits/sec, 0 packets/sec
1244080 packets input, 950518882 bytes, 0 no buffer
Received 1244071 broadcasts (1244051 multicast)
6561 runts, 0 giants, 0 throttles
31176 input errors, 24615 CRC, 0 frame, 0 overrun, 0 ignored
0 input packets with dribble condition detected
874246 packets output, 658148088 bytes, 0 underruns
0 output errors, 0 collisions, 0 interface resets
0 babbles, 0 late collision, 0 deferred
0 lost carrier, 0 no carrier
0 output buffer failures, 0 output buffers swapped out

GigabitEthernet1/5 is up, line protocol is up (connected)
Hardware is Gigabit Ethernet Port, address is 001e.f71a.5944 (bia 001e.f71a.59
44)
MTU 1500 bytes, BW 100000 Kbit, DLY 10 usec,
reliability 255/255, txload 1/255, rxload 12/255
Encapsulation ARPA, loopback not set
Keepalive set (10 sec)
Full-duplex, 100Mb/s, link type is auto, media type is 10/100/1000-TX
input flow-control is off, output flow-control is off
ARP type: ARPA, ARP Timeout 04:00:00
Last input never, output never, output hang never
Last clearing of "show interface" counters 00:36:22
Input queue: 0/2000/0/0 (size/max/drops/flushes); Total output drops: 0
Queueing strategy: fifo
Output queue: 0/40 (size/max)
5 minute input rate 5093000 bits/sec, 839 packets/sec
5 minute output rate 0 bits/sec, 0 packets/sec
1662763 packets input, 1254087837 bytes, 0 no buffer
Received 1662742 broadcasts (1662719 multicast)
339 runts, 0 giants, 0 throttles
2679 input errors, 2340 CRC, 0 frame, 0 overrun, 0 ignored
0 input packets with dribble condition detected
522691 packets output, 393062643 bytes, 0 underruns
0 output errors, 0 collisions, 0 interface resets
0 babbles, 0 late collision, 0 deferred
0 lost carrier, 0 no carrier
0 output buffer failures, 0 output buffers swapped out

Hope to hear from someone soon,

Evan

 

[mamoser]

Here's the switch config that was suggested to me that doesn't seem to work. I see the traffic coming into the g1/1 and 1/5 but my smartbit doesn't see it even though it's sending igmp join messages for the multicast groups.

Switch#sh run
Building configuration...

Current configuration : 2644 bytes
!
version 12.2
no service pad
service timestamps debug uptime
service timestamps log uptime
no service password-encryption
service compress-config
!
hostname Switch
!
boot-start-marker
boot-end-marker
!
!
no aaa new-model
vtp mode transparent
ip subnet-zero
!
ip multicast-routing
!
no file verify auto
spanning-tree mode pvst
spanning-tree extend system-id
power redundancy-mode redundant
!
!
!
vlan internal allocation policy ascending
!
vlan 10
name primary
!
vlan 20
name secondary
!
interface GigabitEthernet1/1 (connected to sat receiver 1)
switchport access vlan 10
speed 100
duplex full
!
interface GigabitEthernet1/2
switchport access vlan 10
!
interface GigabitEthernet1/3
switchport access vlan 10
!
interface GigabitEthernet1/4
switchport access vlan 10
!
interface GigabitEthernet1/5 (connected to sat receiver 2)
switchport access vlan 10
speed 100
duplex full
!
interface GigabitEthernet1/6
switchport access vlan 10
!
interface GigabitEthernet1/7
switchport access vlan 10
!
interface GigabitEthernet1/8
switchport access vlan 10
!
interface GigabitEthernet1/9
switchport access vlan 10
!
interface GigabitEthernet1/10
switchport access vlan 10
!
interface GigabitEthernet1/11 (smartbit)
switchport access vlan 10
!

interface Vlan1
no ip address
!
interface Vlan10
ip address 192.168.10.115 255.255.255.0
ip pim sparse-mode
ip cgmp
!
no ip http server
!
ip pim rp-address 192.168.10.115
ip pim send-rp-announce Vlan10 scope 16 group-list 1
access-list permit 192.168.10.115
!
!
!
line con 0
stopbits 1
line vty 0 4
!
!
end

 

[BuckWeet]

Don't use CGMP as its an outdated protocol. IGMP snooping is the path that cisco is following, CGMP is phased out. Since it looks like that the 4948 is the only device, CGMP isn't even needed.. Just turn on PIM on the interface (which you have done) and IGMP snooping will be enabled automatically.


You do have quite a few interface errors which usually are a sign of speed/duplex issues.

What I'm saying is that you have 4 multicast groups.. Lets take group A for example.. As long as group A's packets are in order why does it matter about the other groups..

Just verify that I stated that correctly as that is how I'm intrepreting it..


Also when you're doing your multicast test could do these commands...

'show ip mroute'
'show igmp snooping groups'
'show proc cpu history'

Also find out what the TTL on the stream is..


BuckWeet

 

[mamoser]

Thanks again for helping me buckweet. I came in today and noticed that someone has taken my modulator. It looks like I won't be getting it back for another day or two. I will get those results ASAP when I get it back.

Sincerely,

Evan

 

[BuckWeet]

DOH!

haha don't you love co-workers

BuckWeet

 

[ADB100]

With regards to CGMP - if the access switch is a 2900XL then CGMP is needed since the 2900/3500XL's don't support IGMP snooping and can only do CGMP. If you are still using a 2900XL then there is something wrong anyway.....

Andy

 

[mamoser]

Sorry guys for not posting in a while. I'm very pissed that I haven't received my equipment back but priority projects override my authority.

This setup has been put on hold until some of the high priority ones are finished. I just want to thank all of you that help and hope that we can continue this in the next week or so.

Evan

 

[burtsbees]

Thanks for the follow-up...keep us posted.

Burt

 

[mamoser]

Hi everyone,

Sorry for taking so long but I still have not received my 4948 switch. I've decided to continue this posting by writing the first part of my setup.

I have some basic questions regarding multicast and was hoping someone here can help me out. As you know, I'm trying to setup a large multicast network for IPTV and audio broadcasting. I've decided to break down the setup in small junks and hopefully get as much help\info\feedback as possible on each section.

I have a large background in networking but unfortunetly I've never really worked with multicasting. I've been reading up on it for the last month but I get confused with the information overload.

My first question is
"How do I control multicating at L2 without a router?"

After reading a lot on this subject, I've come to the conclusion that I need to make static mac table entries. IGMP snooping\CGMP from what i gather requires a mrouter to receive the forwarding table created by the l2 switch.

Is this true?

The first part of the design does not have a router. The video feeds are flooded onto a 2900 sw and are sent to an encapsulator on port fa0/24. Ports fa0/1-8 contain different encoders pumping out seperate multicast traffic.

[Video Encoders]
|||||||| -) ports fa0\1-8
[Cisco 2960 or 2950]
| -)port fa0/24
[Encapsulator]

My multicast address are the following:
230.0.0.10 = 0100.5e00.000A
230.0.0.20 = 0100.5e00.0014
230.0.0.30 = ....
230.0.0.40 = ....
230.0.0.50 = ....
230.0.0.60 = ....
230.0.0.70 = ....
230.0.0.80 = ....

So following the static mac entries path, this is what I was planning to configure on my 2900 sw:

2900(config)#no ip igmp snooping vlan 1
2900(config)#mac-address-table static 0100.5e00.000A vlan 1 interface fa 0/24
2900(config)#mac-address-table static 0100.5e00.0014 vlan 1 interface fa 0/24
2900(config)#mac-address-table ... (rest of multicast adds)

By applying these commands, all multicast traffic from 230.0.0.10-80 will now only be forwarded to fa0/24 but all other multicast traffic will continue to flow.

Does this make sense or am I going at it the wrong way? If i need to implement a l3 sw, I will!

Hope to hear from someone soon,

Evan

 

[mamoser]

Instead of having information all over the place, I've decided to move the thread to the following link:

http://www.tek-tips.com/viewthread.cfm?qid=1476026&page=1

 

[burtsbees]

Does it matter that the acl
access-list permit 192.168.10.115
does not have a number, since
ip pim send-rp-announce Vlan10 scope 16 group-list 1
points to acl 1?

Burt