We have a loaned PS6000E SAN that we are using to relocate our iSCSI LUNs off our current PS4000E SAN to upgrade it from Seagate Moose drives under a proactive replacement plan by Dell. We've connected it up to our stack of Dell PowerConnect 6248 switches that make up our client stack (because our Cisco stack is currently full – another is on order so we can extend it and give our server network room to grow) and while it connects to our existing stack and we've successfully moved a LUN across, there are numerous errors on the Ethernet connectors that are giving us cause for concern.
Our other two SANs are connected to our Cisco stack and show no errors at all. They connect to the Dell stack via a fibre link which does not have jumbo frames turned on.
We are not particularly keen on moving across the remainder of the LUNs until we resolve this issue. Is there something on the switch we need to configure that I'm missing? Should I not be using the PC6248s for this and should I wait for our new Cisco switch to arrive? Am I getting myself into a fuss over nothing?
EDIT: I noticed spanning tree was turned on for the ports and turned off for the SAN ports on the Cisco switch. Turning it off has done nothing to stem the errors coming from this thing. I ensured that the MTU was 9000 across the board and it hasn't made a difference. All cables are CAT6. Turning off LLDP made no difference. Someone requested a diagram, which is below.
Best Answer
Just follow the configuration guide for the PC6248 on the "Rapid Equallogic Configuration Portal by SIS" page: http://en.community.dell.com/techcenter/storage/w/wiki/3615.rapid-equallogic-configuration-portal-by-sis.aspx Also, you're correct to have turned off spanning tree on the array-side ports, this is a common cause of network ports.
These switches are validated for use with Equallogic arrays, so you shouldn't see problems with them, especially on a single array. Keep in mind though, if you're running EQL FW v6.0+ you need a minimum of FW v3.3.4.1 on the switches.
Somewhat. It really depends on what type of errors these are, and how many packet errors there are in comparison to normal/good traffic (i.e. what percentage of traffic is marked as errors).
You might want to try disabling DCB on the Equallogic array to see if that happens to clear the errors up too.
I would highly recommend creating a support ticket to have this looked into though. This is most easily done on eqlsupport.dell.com > Account Management > Log a Case