02-17-2010 06:32 AM
It has been recommended that I set up ISL on our 200e brocade switches. Can someone point me in the right direction to get started on this?
Both switches are v 5.3.0. This has been recommended to provide increased redundancy between our filers. Currently we have two controllers and each on connects to a separate switch.
02-18-2010 08:25 AM
I dont understand the scenario here - how many switches do you have at this point in time??
The precautions prior to creating fabrics (by joining ISL's) is provided on this forum umpteen times. Kindly look it up and you are welcome to ask doubts.
02-18-2010 06:24 PM
if you have two 200e's with your storage arrays controllers going to each switch , the you have two seperate "fabrics".
By joing them together with an ISL you are creating one single fabric , which then does not give you your "redundancy" or "high avilability"
02-22-2010 11:59 AM
@bijukrishnan: I have two brocade 200e switches.
@emilio: the switches are not our core switches but connect our production vmware environment
To be a little more specific I have a VMware environment on FC protocol. Two ESX hosts with two single-port HBAs in each, connected to separate fiber switches. I have attached a simple Visio diagram to better illustrate the setup. Is there something, like ISL I can implement to help make the infrastructure a little more redundant?
02-22-2010 12:33 PM
It looks like your 200E are setup as stand alone switches. To me that makes them core switches . One would be Fabric A and the other Fabric B, The ESX servers are connected to both ( as an example, HBA0 goes to one switch & HBA1 goes to the other ). That seems fine. Your issue is the SAN STORAGE.
1) Your NETAPP Appliance is only connected to one side (switch)
2) Your SAN 2 is connected to the other side (switch)
Make sure these both have 2 controllers, if so then you would look at making a fiber connection to the other switch. What your diagram shows is the ESX servers having 1 connection to each switch which is fine but your storage is lacking redundancy. ie: If that fiber link goes down on the NETAPP you will lose connectivity and there isn't a redundant fiber connection to fall back on.
02-22-2010 01:17 PM
Thanks for the feedback. It would help if I clarified that the two FAS2020's are indeed controllers in the same filer. For ease of visual I separated them. Currently the e0a interface on each connects to their own switch and the e0b connects to them to the shelf.
02-23-2010 01:13 AM
The connection as per vsd file Esx host1 & Esx host2 with single port hba*2 nos on both Esx host is connected to 2 different fabrics which is ok.
There is only one connection from netapp filer FAS2020 to swicth for 2 filers, As per the netapp product guide there are single & dual controller config available, If the Filer has two contorller it should have 4 FC ports, 2 FC ports on each of the Controller.
The ports which you have mentioned E0B & E0B is a network port for Ip connection.. If its typo error then its ok or else if you have shown the 2 controllers right & left for ease of understanding then there should be no change.
Basically as per your Design there should be no ISL link from switch to switch,