|
Home > Archive > Data Storage > December 2005 > Trunking 1620 McData storage routerS for ethernet performance (iSCSI env.)
You are viewing an archived Text-only version of the thread.
To view this thread in it's original format and/or if you want to reply to
this thread please [click here]
| Author |
Trunking 1620 McData storage routerS for ethernet performance (iSCSI env.)
|
|
| goce.delcev@gmail.com 2005-12-24, 5:50 pm |
| Hi group,
this is my first post on this comunity.
I have a question for which cannot find answer neither on google nor
documentation.
I have more 1620 McData storage routers (04M-R IBM nomenclature).
Now, 2 of them should be connected in failover fashion.
1 eth. (iSCSI) port to 1 IP switch.
A lot of iSCSI servers attached to an ethernet swithc.
The problem is that all I have left is 1Gb ethernet port and an
ethernet troughput becomes a bottleneck.
Is there any way to connect 2 1620 McData storage routers to work
together i.e. double the ethernet (very important) and FC troughput?
Any other idea?
Looknig forward to hear from you,
GD
| |
| robprzy 2005-12-26, 5:47 pm |
| Hello,
I am very familiar with the IPS line from Nishan/McDATA. There really
isn't a way to aggregate the switches together in a scalable fashion.
Since you are operating in an iSCSI environment, it really won't buy
you much unless your initiator has multi-path intelligence. Your iSCSI
initiator only has 1 path to your volume through the 1620 gateway. Now
if your iSCSI initiator could log into multiple portals and view a
single lun via multple connections, then you could round robin your IO
amongst the available paths.
Rob
|
|
|
|
|