IGMP snooping

Hello to all!
I have come a cross one problem with hope someone can help me solve it or at least give some valuable ideas.
The problem is regarding IGMP snooping with Cisco 4948E swithes. By documentation it is said that IGMP snooping is turned on by default.
However, I was getting all multicast traffic on the port where there is only one receiver, tuned to only one multicast group.
I was able to solve to problem by turning igmp snooping querier option and now it is working ok.
However, I have to plug in a Cisco router (7200 series). In that case the router is the querier.
My problem is that the router is constantly getting all multicast traffic, causing the CPU time to increase and be crushed.
I have several multicast groups on the switch, but want only specific ones to be present on the port where router is connected.
I think IGMP snooping is not working on 4948E switches.
Did anybody else experience the same problem?
The router has PIM enabled and I have managed to solve the problem with access list being put in outside direction but this is not the neat solution.
How can I solve this problem? When I do show int command on the router I am getting all multicast traffic and I want to receive only specific ones with way too lower traffic amount.
Thank you very much!

Similar Messages

Ip igmp snooping querier on Nexus, what source IP address to use?

Am looking at a problem with servers in the same vlan across multiple switches that are unable to communicate using multicast. I have found that in the systen I'm to set up I need to apply the ip igmp snooping querier command, in the vlan, but it needs a source IP address.
Different documents make conflicting recommendations for this address, one suggests that any unused address will do, another suggests to use the IP address that is configured on the SVI for the vlan.
Which is correct?

Eventually I had to ask Cisco TAC, the response was that any IP address within the subnet could be used. The recommendation was to allocate an unused address in the vlan subnet for this purpose, use the same address on multiple switches should resiliance be required.

Does the Catalyst 2960x-48TD support Cisco's IGMP Snooping MIB?

Does the Catalyst 2960x-48TD support Cisco's IGMP Snooping MIB? We have a number of switches from different vendors and have recently begun switching to Cisco's, however unlike the other brands I am unable to audit the IGMP Snooping settings via SNMP. It appears that while the switch supports IGMP Snooping it does not support the CISCO MIB for IGMP Snooping. Is this a known bug and is it fixed in any of the later releases?
Our company is purchasing and deploying C2960X-48TD's and I am tasked with updating our system audit software

Hi Sean,
Nov 21 15:40:42.390: %MCAST-SPSTBY-6-ETRACK_STATS_LIMIT_EXCEEDED: Number of entries in IGMP snooping explicit-tracking statistics has exceeded the maximum limit (32000)
The above log message is written because the number of explicit-tracking statistics entries is bounded to avoid monopolizing of system resources by IGMP/MLD snooping. The explicit-tracking statistics database maximum size is set to the same as that of explicit-tracking limit.
The statistics are split into two banks: permanent and volatile. The statistics entries will be stored permanently (volatile) until the permanent entries threshold is reached, after which the statistics will be stored temporarily up to the database maximum size. When the size of the database exceeds the permanent threshold, a group will be removed on receiving an IGMP/MLD leave. Groups can be removed whether they were installed above or below the permanent threshold. If the number of statistics entries grows beyond the maximum size of the database, then no statistics will be stored for any newly formed groups.
You can disable ip igmp snooping explicit-tracking by the following
1) Configure "no ip igmp snooping explicit-tracking"
2) Configure "ip igmp snooping limit track 0" on global.
3) After ET is disabled, please do a "clear ip igmp snooping statistics".
Here is a reference URL (please make sure you copy and paste the entire link to the browser):
http://www.cisco.com/en/US/docs/switches/lan/catalyst6500/ios/12.2SX/configuration/guide/snooigmp.html#wp1062242
HTH
Luke

IGMP Snooping IPTV SG200-08

For having 3 settopboxes for iptv I have setup vlan 3 for iptv.
IGMP snooping is enabled for vlan 3 with the default settings and port 5 (incoming iptv) is enabled as IGMP Mrouter.
In the multicast forwarding table vlan 3 is set to filter unregistered.
This is working fine for a while but after 2 to 3 days the settopbox doesn't receive the iptv signal anymore and I have to restart the switch.
After the reset everything is working fine again for a few days and then the above repeats it self.
How can I solve this problem?
Sent from Cisco Technical Support iPad App

Hi Alexandre, I think the only way to see what is happening is by generating a packet capture to see why when one of the connections leave, what is being sent or received by the other to cause the delay.
-Tom
Please mark answered for helpful posts

RedHat Enterprise Cluster and Cisco IGMP Snooping/Querying

Has anyone else had any experience with IGMP Snooping/Querying and RedHat Enterprise Cluster?
We have been experiencing a large amount of problems with this functionality.
We are running IGMP Querying in our environment and we recently set up a second querier.
Here's the steps we took
Existing querier: 192.168.3.248
Everything was running fine.
Added a new querier on a different switch: 192.168.3.247
At this point, all of our RedHat Enterprise Clusters fenced themselves and needed to be restarted in order to restore
access. In order to restart the RedHat Enterprise Clusters, the physical servers must be rebooted.
Are there any known issues with RedHat Enterprise Clustering and Cisco Switches (3750
series)? I would expect the querier change to be seamless, but it does not seem that this
is the case.

Hi,
In our organizaiton we have Red Hat Cluster with 2 cisco switch (Model: cisco WS-C2960S-24TD-L, Version: "flash:/ c2960s-universalk9-mz.122-55.SE3/c2960s-universalk9-mz.122-55.SE3.bin").
- We are using HP Chassis c7000 and Server is on the chassis. There are 2 service IC & Med. Each server has one service primary and other secondary running.
- The two cluster switches are connected each other with Ether channer trunk (1+1) link. Also these 2 switches are connected to our Mgmt switch for Server Admin access to HP Chassis via OA port. The Red Hat system has cluster lan (pri & sec) & OA lan (01 & 02 of HP chassis) connected to Cluster switches. The Mgmt VLAN is 501 - 172.31.10.0/24.
Problem:
When the CluserSW01 goes down the cluser shifted to CluseterSW02 with Cluser_Secondary_LAN and OA2. But when the ClusterSW01 switch comes again than the communication breaks and cluster don come up.
I was thinking this is either STP or IGMP, well sure though. As these are production systems hence we also couldn't do much more test as well.
If you have face any such issue or have experience with it or know what the problem might be... kindly share with me.
Thanks,
Adnan

Ip igmp snooping querier?

i've been reading these forums and noticed that lot of people have problems with configuring multicast. basicly, problems are flooding bandwidth, no querier, confusing pim router configuration ...
also, ive seen some promisses from tehnical stuff that next gen of linksys switches will have #ip igmp snooping querier, like cat ones have. so i'm interested if someone can give me the answer.
it is interesting question for many people who would cant afford expensive cat switches for such simple usage.
regards,
david

Hi David,
You actually have a lot more options as seen in the GUI, so I though I would put as simple network together and run you through a validation.
But I am assuming you have the refreshed version of the SRW224G4P switch called the SRW224G4P-K9-NA .
If you have the old switch the following is not relevant to you.
I put together a simple network diagram showing my simple network topology used during my wee test;
I chose the RV042 as my Layer 3 router in this network.
It really was to provide DHCP services, but shortly after all devices had IP addresses, I removed the RV042, becauise I noticed it was sending out IP IGMP general queries.    Something I didn;t want in the following test.
I was using a software based multicast tester, produced by the wonderful folk at Twisted Pair Solutions. This software can either send and/or receive multicasts of a defined group and responds appropriately to IGMP General queries. I will call this software MCT (Multicast test).
I had the MCT software on PC C. PC C is plugged into switch port 10 It is sending multicasts into the switch at a rate of about 10 multicast packets a second.
I had the MCT software on PC A. PC A is plugged into switch port 1. It is set to receiving and count multicasts from the switch
I had PC B plugged into switch port 2. PC B is checking switch statistics, but not try to receive multicasts from PC 3.
I had setup multicast snooping on VLAN 1 and querying according to the GUI screen capture below.
Note: I only had a single vlan on this switch.
In a composite of two screen captures below, you can see that IGMP querier status is enabled.
I took off the RV042 out of the network, because i noticed within a wireshark capture on PC A,   that the RV042 was sending out group general queries. I wanted the switch to take over that role
I waited a few minutes, and displayed only igmp activity.
It can be clearly seen in the screen capture below;
SG300 series switch was sending out general queries (2 minutes) from source IP address 172.16.1.100.
It waited for the mcast router general query, but when none came, 60 seconds later sent out it's own general query. Fair enough, it had to assume a mcast router might be sending general queries every minute.
Switch statistics can be seen below, it showed;
that PC B on switch port G2 was not being flooded by multicasts, in other words snooping is working;
Port 10, PC C is sending in 2962 multicast packets to the switch.
Port 1,   PC A is receiving 2900 multicast packets from the switch
port 2 ,   PC B is not part of the multicast group 234.55.66.77 and as such is not being flooded by multicasts.
ALL three PCs are windows based products, and certain multicasts emanate from these devices.
show interafce count
      Port       InUcastPkts InMcastPkts InBcastPkts    InOctets
      gi1             12           9            6           3493
      gi2            247          237           6          43733
      gi3             0            0            0            0
      gi4             0            0            0            0
      gi5             0            0            0            0
      gi6             0            0            0            0
      gi7             0            0            0            0
      gi8             0            0            0            0
      gi9             0            0            0            0
      gi10            26          2962          3          850881
      gi11            0            0            0            0
      gi12            0            0            0            0
      gi13            0            0            0            0
      gi14            0            0            0            0
      gi15            0            0            0            0
      gi16            0            0            0            0
      gi17            0            0            0            0
      gi18            0            0            0            0
      gi19            0            0            0            0
      gi20            0            0            0            0
      gi21            0            0            0            0
      gi22            0            0            0            0
      gi23            0            0            0            0
      gi24            0            0            0            0
      gi25            0            0            0            0
      gi26            0            0            0            0
      gi27            0            0            0            0
      gi28            0            0            0            0
      Port       OutUcastPkts OutMcastPkts OutBcastPkts OutOctets
      gi1             13         2900          9          839691
      gi2            233         343           9          90453
      gi3             0            0            0            0
      gi4             0            0            0            0
      gi5             0            0            0            0
      gi6             0            0            0            0
      gi7             0            0            0            0
      gi8             0            0            0            0
      gi9             0            0            0            0
      gi10            24          343           12         47165
      gi11            0            0            0            0
      gi12            0            0            0            0
      gi13            0            0            0            0
      gi14            0            0            0            0
      gi15            0            0            0            0
      gi16            0            0            0            0
      gi17            0            0            0            0
      gi18            0            0            0            0
      gi19            0            0            0            0
      gi20            0            0            0            0
      gi21            0            0            0            0
      gi22            0            0            0            0
      gi23            0            0            0            0
      gi24            0            0            0            0
      gi25            0            0            0            0
      gi26            0            0            0            0
      gi27            0            0            0            0
      gi28            0            0            0            0
If I look at the multicast groups within the switch, it show that only two ports were involved in multicast group 234.55.66.77. This is the correct behaviour.
switch4cf17c#sh ip igmp snooping group
Vlan   Group        Source        Include Ports        Exclude Ports     Comp.
       Address      Address                                              Mode
1    234.55.66.77 *            gi1,gi10                                  v2
1    239.255.255. *            gi1-2,gi10                                v2
     250
Just to complete the posting, here is the CLI command (highlighted) to enable the querier functionality within the switch;
[Kswitch4cf17c#sh run
bridge multicast filtering
ip igmp snooping
ip igmp snooping vlan 1
ip igmp snooping vlan 1 querier
interface vlan 1
ip address 172.16.1.100 255.255.255.0
exit
interface vlan 1
no ip address dhcp
exit
hostname switch4cf17c
no passwords complexity enable
no snmp-server server
ip telnet server
Hope this helps you and others
regards Dave

CSM Fault Tolerance and IGMP Snooping

For "connection redundancy" the redundancy guide says to turn off IGMP snooping.
Is there any way around this?
I need to have multicasting everywhere and I don't want to multicast all streams to every port on this switch.

Most of the show statements are at the end of the attached file in an earlier post. The vservers details are at the end of this post.
I have an ARP entry for the VIP - 0001.64f9.1a64, but it does not respond to pings. I tried both the alias and the server vlan IP as the default gateway of the servers.
I took a trace and found that the VIP sends a TCP reset immediately after a request. I have tried versions 4.2.1 and 4.1.4 with the same result. I wonder if this could be a problem with the Sup720 with 12.2.17d IOS. I also tried the CSM in slots 2 and 3.
720Test2#sh mod csm 3 vserver detail
SOFTRICITY, type = SLB, state = OPERATIONAL, v_index = 10
virtual = 10.10.249.6/32:0 bidir, any, service = NONE, advertise = FALSE
idle = 3600, replicate csrp = none, vlan = ALL, pending = 30, layer 4
max parse len = 2000, persist rebalance = TRUE
ssl sticky offset = 0, length = 32
conns = 0, total conns = 1
Default policy:
server farm = SOFT1, backup =
sticky: timer = 0, subnet = 0.0.0.0, group id = 0
Policy Tot matches Client pkts Server pkts
(default) 1 1 0

Multicast mac address isn't learned, igmp-snooping

I have PIM router which connects to the cat 2960 switch and also I have host which connects to another port on the same switch. Host was joined to the IGMP group 224.1.1.1. I see that the router generates igmp-query and the host respons. IGMP-snooping process sees that process and updates appropriate entries:
2960-5#sh ip igmp snooping mrouter
Vlan ports
15 Gi2/0/32(dynamic)
2960-5#sh ip igmp snooping groups
15 224.1.1.1 igmp v2 Gi2/0/32, Gi2/0/33
But when I command "sh mac address-table multicast" I see nothing:
2960#sh mac address-table multicast
Vlan Mac Address Type Ports
What is reason of this problem?

There is the following statement from the "CCNP Practical Studies: Switching:
the process of populating the bridge table with multicast MAC addresses is based upon inspection of the destination MAC address, unlike unicast MAC addresses where the source MAC address of unicast frames is examined to generate bridge table entries.
And this book describes other parts of the mac learning process and says that after exchanging IGMP-message MAC-table must be populated by multicast mac-addresses. But later I found some Cisco and Jupiner documentation which says there is two way to perform multicast forwarding - MAC and IP. Default metod is IP multicast forwarding. When this metod is used multicast MAC-addresses isn't learnt and process of packet forwarding uses special forwarding cache which includes list of mapping IP and appropriate interfaces. It all means that this book isn't actual. All modern switchs perform multicast forwarding by IP metod and MAC-addresses don't populate CAM.

SRW208: IGMP Snooping not working!

Hi all,
at the moment we have trouble with the following SRW208 setup:
- Port mirroring is enabled (Port 8 mirrors Port 1) for sniffing via Wireshark in promiscuous mode.
- 'IGMP Snooping' is enabled for all ports.
Our test scenario:
- an IPTV settop-box is connected to port 1 and gets/sends traffic from/to a VDSL router connected to port 2.
- a wireshark PC sniffs at mirror port 8 (all traffic) in promiscuous mode
- another wireshark PC sniffs at port 3 (all traffic) in promiscuous mode
With IGMP snooping enabled only the wireshark PC at (mirroring) port 8 should be able to sniff the mcast traffic between the settop-box and the VDSL router.
The problem is that wireshark can sniff the mcast traffic at all ports. In addition the LED of all active ports are blinking synchronously/continuously.
It seems that the IGMP snooping does not to work properly because all ports are still flooded with mcast traffic.
With an old Zyxel switch the same setup works without a problem.
Is this a known problem of the srw208?
Cheers,
Fred

So much for the image upload, how about this diagram:
/* Style Definitions */
table.MsoNormalTable
{mso-style-name:"Table Normal";
mso-tstyle-rowband-size:0;
mso-tstyle-colband-size:0;
mso-style-noshow:yes;
mso-style-priority:99;
mso-style-qformat:yes;
mso-style-parent:"";
mso-padding-alt:0in 5.4pt 0in 5.4pt;
mso-para-margin-top:0in;
mso-para-margin-right:0in;
mso-para-margin-bottom:10.0pt;
mso-para-margin-left:0in;
line-height:115%;
mso-pagination:widow-orphan;
font-size:11.0pt;
font-family:"Calibri","sans-serif";
mso-ascii-font-family:Calibri;
mso-ascii-theme-font:minor-latin;
mso-fareast-font-family:"Times New Roman";
mso-fareast-theme-font:minor-fareast;
mso-hansi-font-family:Calibri;
mso-hansi-theme-font:minor-latin;}
Spirent 3 --------------| SRW2008 |------------------ Spirent 4
                            |                   |------------------ Spirent 5
                            |                   |------------------ Spirent 6
                            |                   |------------------ Spirent 7
                            |                   |------------------ Spirent 8

IGMP Snooping configuration for Multicasting on Cisco Catalyst 3020

IGMP Snooping configuration for Multicasting on Cisco Catalyst 3020
Hi Guru,
Our switch model is "Cisco Catalyst Blade Switch 3020 for HP"
We are building HA (High Availability) Databases infrastructure.
Currently, there are two nodes(hosts- servers) and two above switch for HA.
Oracle said we need to turn off the IGMP Snooping in order to use the multicasting for their interconnect communication.
So my question is:
Q1> Is there any way to use Multicasting without turning off IGMP Snooping on Switch side?
Q2> If 'yes', how can we configure the switch for Multicasting ?
Oracle uses 230.0.1.0 & 224.0.0.251 IPs with 42000 range port for Multicasting communication.
Please shed some light on it

May I request if you avoid making multiple posts of the same issue? This will only confuse people (including you).

ESW 520 IGMP Snooping

Hi all,
I have to deploy a full IP-TV solution and the IGMP snooping must be supported on the switch.
I know the the ESW 520 support this protocol but the full IP-TV solution will manage uo to 200 set up boxes, does this switch support all this traffic?
I also have another question, what exactly mean
"Combo SFP slots include one 10/100/1000BASE-T Ethernet port and 1 SFP-based Gigabit Ethernet slot for fiber, 1 port active at a time." ? Is when I use 1 port with LX sfp transcever the 3 other uplink ports will be inactive and I can't add another or trunk it with another switch?
Thx for the support.
BR

Hello Hakim,
I cannot answer your first question. I think that if you provide some more detail, such as how much traffic 200 "set up boxes" will generate, you may have a better chance of a response.
Regarding your second question, you can have two trunks using EITHER the two 10/100/1000BASE-T Ethernet ports OR the SFP-based Gigabit Ethernet slots for fiber OR one of each. Of the four ports you can only use two at a time because they are "shared".

Catalyst 2955T-12 Power Supply n IGMP Snooping

Hello everyone,
I have Two Catalyst 2955T-12 switches, on this switch there are Two Power Supply (DC) options...I want to implement redundant Power Supply solution, can this Two DC Input on Switch will give this solution or I have also external Din Rail power supply...my question is how I can implement redundant power supply so that if one PS of switch fails then automatically other will start?
Also can any body tell me what IGMP snooping do? do I need to configure it? also the command (I need global command as well as port/interface based command)?
Regards,
Shahid

For the power redundancy question, I believe this link will help:
http://www.cisco.com/univercd/cc/td/doc/product/lan/cat2950/12122ea2/2955hig/hginstal.htm
Also can any body tell me what IGMP snooping do? do I need to configure it? also the command (I need global command as well as port/interface based command)?
>>Layer 2 switches can use IGMP snooping to constrain the flooding of multicast traffic by dynamically configuring Layer 2 interfaces so that multicast traffic is forwarded to only those interfaces associated with IP multicast devices. As the name implies, IGMP snooping requires the LAN switch to snoop on the IGMP transmissions between the host and the router and to keep track of multicast groups and member ports. When the switch receives an IGMP report from a host for a particular multicast group, the switch adds the host port number to the forwarding table entry; when it receives an IGMP Leave Group message from a host, it removes the host port from the table entry. It also periodically deletes entries if it does not receive IGMP membership reports from the multicast clients.
The multicast router sends out periodic IGMP general queries to all VLANs. When IGMP snooping is enabled, the switch responds to the router queries with only one join request per MAC multicast group, and the switch creates one entry per VLAN in the Layer 2 forwarding table for each MAC group from which it receives an IGMP join request. All hosts interested in this multicast traffic send join requests and are added to the forwarding table entry.
IGMP Snooping is by default Enabled globally and per VLAN. If for some reason it is not enabled, "ip igmp snooping" at the config mode to enable it.
There is nothing else to configure.
http://www.cisco.com/univercd/cc/td/doc/product/lan/cat2950/12122ea5/2950scg/swigmp.htm#wp1020034
Please rate helpful posts.

Multicasting (IGMP Snoop) between Nortel and Cisco

We are currently having issues with Zen imaging (multicasting) and our setup is the following.
Please take into account, our knowledge is very limited with IGMP Snooping setup etc.
MDF = 6 Nortel 450-24T's using FirmWare -1.48 / SoftWare - 4.5.2.4
IGMP Settings are such :
VLAN: [ 1 ]
Snooping: [ Enabled ]
Proxy: [ Disabled ] -----> This was on...but once off, runs much smoother.
Robust Value: [ 2 ]
Query Time: [ 125 seconds ]
Set Router Ports: [ Version 1 ]
In the MDF (anythig directly in those switches) images fine now. (once I disabled PROXY)
However I have a few IDF's off the MDF that are using OLD Nortel 350F-HD's (no IGMP Snooping support) and it's horrible (can only do a few computers at a time.
So in one of the IDF's (the biggest one) I pulled out the 350F-HD and replaced it with a CISCO 2950 w/Fiber and it's using 12.1.20EA1 and I left IGMP Snooping on (thinking this will fix it) and couldn't even get ONE machine to connect and image in the multicast session. It's settings were (by default):
Global IGMP Snooping configuration:
IGMP snooping : Disabled
IGMPv3 snooping (minimal) : Enabled
Report suppression : Enabled
TCN solicit query : Disabled
TCN flood query count : 2
Vlan 1:
IGMP snooping : Disabled
Immediate leave : Disabled
Multicast router learning mode : pim-dvmrp
Source only learning age timer : 10
I then completly disabled IGMP Snooping on the CISCO and we're able to Image 5-7 Computers without a crash (more than that and it crashes - disconnects etc)
In the area's that I have All 450's or all CIsco's the imaging seems to go fine. (with minor errors)
Can any one give me some advice (or hopefully ran into this mixed setup before)?
Thank you.

Bosalaza,
Thank you for replying (and I read even more on the ip multicast routing). However I've not ran into the same issue at any school that has 100% cisco switches or 100% Nortels (that are setup correctly and not older than dirt). I think we've not needed the multicast routing setup as we only have one router on the network (and it's flat at the moment anyway). As long as IGMP Snooping is enabled correctly (on the switches) it seems to serve us well.
Although from what I've read (where you pointed me too) it seems even in our setup we would benifeit from taking time to setup "ip pim ....." etc.
I was able to scrounge from another network and change out a few very old Nortels (that didn't support IGMP Snoop) and all seems well now.
So long story short (and incase anyone else needs this info. The Nortel 350T and F - HD's were the main issue. It seems (for now) that a mixture of Nortel 350/450-24T's (any model that at least has IGMP Snooping) and Cisco's mixed (also Snoop on) works pretty well.
I'm going to consider this solved as I was able to fix it with changing out some old product. However I really appreciate your efforts and pointing my towards some good info. (Which I'm going to read up on more, as I'm sure we'll need to get it setup in the near future.)
Thanks again.

IGMP snooping functionality on apple time capsule & extreme

Hi,
I use an apple time capsule which has 7.7.3 version in it. There is an option in apple capsule to enable or disable IGMP snooping,
Does apple capsule disables forwarding multicast packets on disabling IGMP snooping?
On wire-shark, I can see an IGMP query message is going out of apple capsule on enabling "IGMP snooping" and there is no IGMP query message from apple capsule when IGMP snooping is disabled. Based on IGMP V3 standard, a IGMP query is sent by router. On "IGMP snooping" disabled case, is it host's responsibility to forward multicast packets?
And I have an airport extreme as well which does not have IGMP snooping enable/disable functionality. How IGMP and multicasting works on this?
Thanks

And I have an airport extreme as well which does not have IGMP snooping enable/disable functionality. How IGMP and multicasting works on this?
What model and firmware is the extreme?
If you look under the wireless options in the 5.6 utility it shows multicast as preconfigured to on.. and set to low by default. you had the option to set higher if you wished with the result that more of the wireless bandwidth was reserved to handle multicast.
You might need to use the older utility if it is Gen5 or earlier as some functionality was simply removed in the v6 airport utility and never returned.
IGMP never rates a mention.
On wire-shark, I can see an IGMP query message is going out of apple capsule on enabling "IGMP snooping" and there is no IGMP query message from apple capsule when IGMP snooping is disabled. Based on IGMP V3 standard, a IGMP query is sent by router. On "IGMP snooping" disabled case, is it host's responsibility to forward multicast packets?
I cannot answer your specific question.. but is the IGMP set on by default?? If so I expect apple was thinking it should automatically handle the bandwidth demand whereas the older ones had issues with this and needed manual setting when using wireless. If you turn it off then multicast will have issues I am guessing.

IGMP Snooping in Switch Hierarchy

Hello,
I have a question regarding IGMP when more than one switch is between mcast router and the hosts.
My setup is as follows:
mcastrouter1
|
Core1
| \
| Access-Switches (several) - hosts
| /
Core2
|
mcastrouter2
The core-switches are cat6k (native), the access-switches a mix of cat3550 and cat4506.
The mcast routers are connected to the core switches, the hosts connect to the access switches. IGMP Snooping is enabled on all switches.
On the core switches I see the IGMP entries in the MAC table just fine, but they somehow seem not to arrive at the access-switches. We tried all the debugs we can think of, but we cannot see any IGMP messages arriving at the access switches, where the hosts are connected. The mcast routers and the hosts are in the same VLAN. This VLAN is configured on the core and the access switches. One of the core switches is the STP root.
The trunks between the core and the access switches are dot1q trunks over glas links. All switches use 12.1 software.
Is this behaviour works-as-designed? I would expect that the IGMP messages are flooded throughout the VLAN, and this is terminated on the access switches. So the access switches should see the messages as well. Our tests tell us that the IGMP messages do not make it to the access switches. I cannot see why the switch should filter the message, because it should be forwarded to the hosts in any case.
Every document I checked on CCO has always one switch between the mcast router and the host.
If anyone has any tips or documents that may help me, I would really appreciate it.
Thanks
-birgit

Hi,
Not sure if this helps much but we are using CGMP in our environment as we have lots of legacy switches at the moment.
I can see lots of CGMP join messages throughout the layer2 switches and would expect to see the same for IGMP
2005 Aug 04 09:27:09 GMT +00:00 %MCAST-6-RX_CGMPJN:CGMP: Rcvd CGMP Join Mesg vla
n 20
2005 Aug 04 09:27:09 GMT +00:00 %MCAST-6-RX_CGMPJN:CGMP: Rcvd CGMP Join Mesg vla
n 40
2005 Aug 04 09:27:09 GMT +00:00 %MCAST-6-RX_CGMPJN:CGMP: Rcvd CGMP Join Mesg vla
n 40
2005 Aug 04 09:27:10 GMT +00:00 %MCAST-6-RX_CGMPJN:CGMP: Rcvd CGMP Join Mesg vla
I'm looking for a switch that supports IGMP so i can enable snooping and see what happens and will report back.
Thanks
Paddy

IGMP snooping

Similar Messages

Maybe you are looking for