[j-nsp] EX4200 VC odd routing issues

Bill Blackford BBlackford at nwresd.k12.or.us
Mon Aug 9 15:52:36 EDT 2010


Thanks Dan,

I believe there's a strong likelihood that 10.0S6.1 may solve my MCAST issue.

I'm still gathering fact of the routing issue and my testing appears to be inconsistent with the system's folks testing.

On a related note, what are the recommendations and config options for EX's where system cluster nodes are distributed across cabinets and different switches (stand alone, not VC),  MAC aging, grat. ARP, etc. ?

Thanks,


-b



-----Original Message-----
From: Dan Farrell [mailto:danno at appliedi.net] 
Sent: Monday, August 09, 2010 12:19 PM
To: Bill Blackford; 'juniper-nsp'
Subject: RE: EX4200 VC odd routing issues

You are encouraged to take your 4200 VC and upgrade it to 10.0S6.1. I realize that you are likely not using hundreds of RVI's, but if you are, this similar behavior can be exhibited. This is what happened- I had about 750 RVI's on a 4200 VC, and it was like a overfilled tomato truck of how many could actually be working at any given time (as one tomato gets on the front of the truck, one would fall off the back), and would randomly not work with some at random times.

It was narrowed down to an issue specifically addressed in 10.0S1.1, now 10.0S6.1. There are other PR's mentioned that seem to touch protocols you're having issues with, you should review this link and use it-

https://www.juniper.net/alerts/viewalert.jsp?actionBtn=Search&txtAlertNumber=PSN-2010-06-820&viewMode=view


 If you are interested in the details of my specific case I'll be happy to share them.

Dan Farrell
Applied Innovations Corp.
danno at appliedi.net


-----Original Message-----
From: juniper-nsp-bounces at puck.nether.net [mailto:juniper-nsp-bounces at puck.nether.net] On Behalf Of Bill Blackford
Sent: Monday, August 09, 2010 2:06 PM
To: 'juniper-nsp'
Subject: [j-nsp] EX4200 VC odd routing issues

EX4200-24T X2 (VC) Layer3
10.1R1.8

EX3200-48T (switch1 and switch2) Layer2 rack switches.
10.0S1.1



I don't have enough empirical data to offer but I figured I'd ask the group in case any of this behavior has been seen.

Problem 1. Inter-vlan routing.
	I'm seeing issues with simple routing from vlanA on switch1 to vlanB on switch2. Both switch1 and switch2 connect via L2 ae trunks to the EX4200 VC. All Vlans are consistent throughout all switches. RVI's on the Layer3 VC are correct. I am also seeing a traceroute from hostA on switch1 to hostB on switch2 to return only the dest host, not the gateway (VC) followed by the dest host as would be expected (IOW, a single hop). The EX4200 just replaced a C3750 VC, the rack switches mentioned did not change.




Problem 2. (separate issue) Multicast on same switch/same vlan.
	I have two hosts that can normally communicate via mcast that now that they're on a Juniper EX cannot. I haven't changed the default of "igmp-snooping vlan all". I move these hosts back to the "C" switch and mcast functions again. 


Thank you in advance for any shred thoughts.

-b

--
Bill Blackford                     
Senior Network Engineer            
Technology Systems Group           
Northwest Regional ESD             

Logged into reality and abusing my sudo priviledges


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



More information about the juniper-nsp mailing list