We have 2 cat 6500 in vss mode, these are the access switches to our storage systems. Here there are 2 models of storage attached to the switches in redundant scheme: emc vnx 5500 and emc vnx 5700. Each storage has two controllers one attached to VSS-A and the other to VSS-B. The problem is that only 5500 storage systems are pingable when both controllers are attached to the switches and 5700 controllers are not pingable or sometime one is pingable and quickly time out happens again.
To figure out the problem I connected all of controllers and my laptop to a 2960 switch and every thing is fine but as soon as I connect the switch to one of the 6500 switches, the same problem happens again.
The other test: I connected all controllers to the line card installed in VSS-A chassis, then the problem solved. Same test for VSS-B chassis.
I checked the mac address table on 6500 switches and mac addresses were there for all controllers but the arp entry for 5700 controllers were suddenly cleared out.
I posted the same issue in Cisco support forum and people there suggested it may be an incompatibility between 5700 and VSS which seems odd to me because storage systems does not understand VSS, they just see one single switch in front of them.
It's really confusing, I don't know what to check any more, any suggestions?
The VNX (5500 and 5700) have two Storage Processors (SPA and SPB - I think you're called these controllers). Each SP has a LAN Management interface using a IP interface that supports 10/100/1000Mb speed. The default on the SP's is set to auto/auto - set the switch ports to auto/auto and see if that helps. Make sure the SP Management Port is also set to Auto/Auto and not set to a specific speed.
From experience to me a compatibility problem seems unlikely – I would first check for config problems on both the VNX and the Cisco side – like wrong VLAN, …