1847067 Members
5339 Online
110261 Solutions
New Discussion

HP9000 MLT/CEF

 
Derek McCutcheon
New Member

HP9000 MLT/CEF

I've read the paper on Cisco CEF, this is a cisco proprietry protocol , does anyone know if there is a similar doc relating to MLT (Multi Link Trunking) which is the Industry std. We have a cust who has a HP9000 cluster with a virtual IP, packets are being intermittantly dropped, the cluster is linked to 2 seperate NOrtel Passport 8600 switches in an MLT configuration, we can see all IP adress in the arp table of both switches ( the real & virt IP address )but the fec doc would indicate that only the virt ip should be seen by both switches. any help / pointers ???

Tnx
Derek
2 REPLIES 2
Mike Hassell
Respected Contributor

Re: HP9000 MLT/CEF

Derek,

I've seen Cisco's "Etherchannel" in action and have been pleased with the results, however I have never worked with MLT on Nortel gear. Anyway, I came across a few things that may be of assistance to you.

1. A breif overview of networking concepts that touches on MLT (Page 3-17)

http://netlab.fe.up.pt/documentacao/nortel/accellarRouting.pdf

2. CLI switch management from nortel that again touches on MLT (The link is long so it may cause problems):

http://www.nortelsupport.co.kr/ntp/data/Passport_8100/Reference%20for%20the%20Passport%208000%20Series%20Command%20Line%20Interface%20Base%20Switch%20Management%20-%20Release%203.1.pdf

3. A good Network World Fusion review of trunking in general, talks about Nortel gear and how they stack up to Cisco when dealing with trunking:

http://www.nwfusion.com/reviews/2001/0416rev2.html

Unfortunately I didn't come across a good whitepaper on the subject as it looks like Nortel requires a login to obtain more detailed information on the use of MLT. Hope that helps a bit.

-Mike
The network is the computer, yeah I stole it from Sun, so what?
Derek McCutcheon
New Member

Re: HP9000 MLT/CEF

Thanks for the reply , yeah i've read both the NOrtel pages before ( Nortel employee !) i was working a customer fault with 2 x 2 sets of clusters which both had similar probs so i was just wondering if there was anything specific in the HP configs, the fault seemed to be with the different nodes within the cluster having differing VRRP addresses instead of the virtual address when we set it to the virtual address the problem went away, the only reason i was asking about trunking is because of the varied & wonderful problems I have seen/experienced i expected that to be the fault not VRRP (HSRP in cisco speak) , thanks anyway