Hp to brocade link aggregation

[RAHUL55]

Hai,

I have some communication  problem with Hp 2610 to Brocade 6610 connectivity, here is my scenario

i have 10 edge switches of HP 2610 and 2 core switches of brocade 6610f  for redundancy and a routing switch of hp 6200 yl, here am using two different network ie 10.246..127.1/24 & 10.246.128.1/24. All the edge switches have dual fiber up link and one up link is go to core sw 1 and second will go to core sw 2, from each core switch one uplink is connected to router switch. All the network is fine when single uplink from each edge switch is connected to core sw-1, when i connect the second uplink to second core switch some loop is happening and network became down, can someone help me for a good configuration

[parnassus]

How your two Brocade ICX6610-24F switches are inter-connected together physically (links: standard 40 Gbps QSFP ports?) and logically (configuration) at core level? Are they stacked (forming a single logical Virtual Switch made of Master and Standby units)?

I'm not an expert but discovering how the core was setup can help diagnose what (and why) is happening between your core level and your edge level when you link any of your edge switches (using a second uplink) to the second core unit.

Also knowing how those uplinks were configured (a Single LAG made of two links? two separate Trunks each one of one link? A Distributed-Trunk?) would be of help.

[RAHUL55]

Hi

Thanks for your reply..Actually both the core switches are independant swtches there is no stacking between the core switches and there is no logical configuration also .

At the edge level also not configured any lag or trunk. So which is the recommonded configuration at edge level and core level.

please  give me solution for configuration at edge level and core level.

[parnassus]

---

RAHUL55 wrote: Actually both the core switches are independant swtches there is no stacking between the core switches and there is no logical configuration also.

You initially wrote "i have 10 edge switches of HP 2610 and 2 core switches of brocade 6610f  for redundancy and a routing switch of hp 6200 yl" so, as I understood it, your two Brocade ICX6610-24F switches placed at core are necessarily interconnected in some way (maybe they are not directly interconnected together but for sure they are indirectly through edge and through the upstream HP 6200 yl switch) and they necessarily have a configuration that provide that redudancy protection...otherwise in which way they provide the redudancy you're claiming for? Is it possible that you have VRRP configured at core level and your two Brocade switches, acting as real core ones, are configured to provide Layer 3 functionalities to downstream devices (edge switches)?

Difficult to say what is happening (apart ascertain that probably a mechanism like STP is intervening because a Loop between your edge and your core was detected when you link those layers with a second uplink per each edge Swtich) if you are telling us there is nothing more than that you already described about your scenario: IMHO your description of your network topology is far to be complete or, if it is really as you wrote, there isn't redudancy at core level...or, more, what you call "core" isn't acting as the "core".

I can't help more far of that point...sorry.

[Vince-Whirlwind]

Enable spanning-tree on all switches.

Set the 6200 to have priority 0.

Set 1x 6610 to have priority 1 (4k)

Set the second 6610 to have priority 2 (8k)

Leave all your Access 2610s with the default priority 8 (32k)