07-11-2008 02:18 AM
I'm planning to cascade my two Brocade 200E switches. Are there any steps to follow regarding this?
Currently this two switches are not yet production though they already have some configurations defined, but are there any considerations to take before configuring those for ISL?
'Appreciate your help. Thanks!
07-11-2008 04:31 AM
--->>> I'm planning to cascade....
for Failover or to expand the Fabric ?
in case of Failover the both switch must have the same config.
Both switch must have a unique Domain ID.
07-13-2008 01:14 AM
Thanks for the reply.
My intention is to expand the switch but at the same time do a Fail Over since the two switches is initially intended for multi-path MPXIO.
I just found out that having a Active-Passive type of Storage needs a direct access of either of its controller's ports when a volume is shared to more than 1 hosts.
Is it possible?
07-13-2008 09:25 AM
yes this is possible.
Almoust all Storage Array support Multi-Path MPIO.
You must also use a Storage Array that Support this option, example by EMC is this called PowerPath.
07-13-2008 06:32 PM
yes this is possible.
Just for clarification, when you said it's possible, do you mean it's possible that the two 200E cascaded switches can work as "Expanded" and "Fail-over" switch?
As you have previously mentioned that for a switch to do a Fail-Over, both switch must have the same config (meaning two independent Effective config).
As I am also aware that an Expanded Switch should basically has only one Effective config. How can I let the switches work as Expanded fabric and at the same time to do Fail-Over?
07-13-2008 10:58 PM
--->>> ...both switch must have the same config...
Both switch must have same FabOS release and the same switch config ( Fabric Parameter ) excepted the Domain ID, each switch must have a unique Domain ID.
switch1 DomainID 1 and switch2 DomainID 2.
Both switch must have same PID Mode, standard is PID 1.
In case of Full Failover, connect both switch as ISL, all Server and Array must be connect on both switch, this mean each server must have 2 HBA, HBA1 is connected to switch1 and HBA2 to switch2, the same by the Storage Array, each Array Port must be Connect to independed Switch.
By the Config, also Zoning, Trunking, etc.... it is not necessary to create this on both switch, it is enough you create on the one of both 1 or 2, after the config is created and enabled, this will automatically Sync in the other switch, all Zoning Port or WWN Zoning are visible on Both switch in the same Fabric.
--->>>How can I let the switches work as Expanded fabric and at the same time to do Fail-Over?
Or is one of both the Switch a this time in Productive Environment's ?
07-14-2008 06:30 PM
Thanks for the helpful information. Both switches are still not in production environment although the Storage that they are currently connected is in production because other environments are also using the storage.
I was able to cascade the two switches and both of them can see only one effective zone configuration as what you have mentioned.
The effective zone configuration involves all the WNNs of all the initiators across the switches together with its storage ports.
Each host with dual-ported HBA will be connected evenly across the two cascaded switch for multipath MPXIO setup.
The Storage that I am using is an Active - Passive type.
If both switch can see only one effective configuration, and in the event of 1 Switch failure, does the MPXIO can support and prevent downtime?
07-15-2008 09:06 AM
if both switch are in the same Fabric ( Failover ) and both have the same config, this is like as MPIO.
You must make sure that the server HBA and the Array is MPIO correct configured.