NETGEAR is aware of a growing number of phone and online scams. To learn how to stay safe click here.
Forum Discussion
rgubele
Apr 28, 2020Aspirant
IGMP Snooping filtering OSPF except on VLAN 1
Hello, I am experiencing an issue where OSPF routers connectted m4300-series switches running IGMP Snooping are unable to form an adjacency. The m4300-series switches seem to be filtering OSPF mu...
- Apr 30, 2020
In 12.0.11.x we introduced a new command 'set igmp-plus" which is needed for all reseved multicast addresses to work.
I beliveve following configuration should fix your issue
vlan database
vlan 20
set igmp-plus 20
rgubele
Apr 29, 2020Aspirant
Hi kevin_hong ,
Thank you for your reponse. My architecture is very complex at the moment, but a simplified version would be something like:
[ROUTER] <-> [52G/24X stack] <-> [12x12f] <-> [ROUTER]
Router could be:
- A Cisco device
- A Mikrotik device
- A Linux machine running Quagga (old and new versions)
- A Juniper device
- An Extreme/Brodcade device
Whether a particular port is trunk or access, whether the vlan is tagged or not, doesn't seem to matter.
Some notable differences I see in our configuration:
We're not using IGMP-Plus mode. When I looked at it, it didn't appear to be appropriate for our configuration because we use any-source multicast and IGMPv2. IGMP Plus mode enables report flooding, which in theory should activate host report suppression and wouldn't work. Additionally, I have Exclude Mrouter Interface Mode disabled, and as nice as fast leave auto detection would be, I've disabled it because it doesn't seem to be reliable and I only need fast leave on a very small number of select reports.
Also, we are using the switches as switches. I'm pretty sure we don't have any layer 3 features turned on, save for one routing interface we use as a management and monitoring IP. It's possible that using OSPF on the switch itself works and doesn't trigger this problem; I don't know.
kevin_hong
Apr 30, 2020Apprentice
Hi rgubele ,
Thanks for your informations.
I think we can do a simple test to verify the cause of the issue.
Case#1: M4300-M4300 [52G/24X stack] <-> [12x12f] --- I've verified that it works
Case#2: Cisco-M4300 --- Could you please verify this case ? (Routing vlan or IP routing interface)
BTW:
For M4300 the MTU was 9198 by default, So It affects protocol message processing with third party devices.
I suggest you try changing the MTU to 1500 on M4300.(Make the MTU of both devices the same)
For VLAN:
(M4300-96X) (Config)#interface vlan 20
(M4300-96X) (Interface vlan 20)#ip mtu 1500
For Interface:
(M4300-96X) (Config)#interface 1/1/1
(M4300-96X) (Interface 1/1/1)#mtu 1500
Thank you.
- rgubeleApr 30, 2020Aspirant
Hi kevin_hong ,
The routing table in my environment is too large to enable OSPF on the m4300s, and I don't really have enough spare equipment to lab it out.
I can confirm that Cisco <-> 4300 <-> Cisco doesn't work with IGMP Snooping enabled, as I observed this in my production environment.
- sdesigowdaApr 30, 2020Initiate
In 12.0.11.x we introduced a new command 'set igmp-plus" which is needed for all reseved multicast addresses to work.
I beliveve following configuration should fix your issue
vlan database
vlan 20
set igmp-plus 20- rgubeleApr 30, 2020Aspirant
Hi sdesigowda ,
Thank you for your thoughts. Netgear support just said essentially the same thing to me: Getting these entries now requires igmp-plus mode.
As I mentioned earlier, I did not enable igmp-plus mode for our environment because it appears that it enables features that are undesirable. I have written back to Netgear support to clarify if turning it on will break my environment or not, and I'm working on the maintenance window to give it a try tonight either way.
Related Content
NETGEAR Academy
Boost your skills with the Netgear Academy - Get trained, certified and stay ahead with the latest Netgear technology!
Join Us!