I've never tried this, but from my experience with other SAN storage devices on other Unix OSs I can give you some pointers.
It's best not to think about installing the DS4300 on Linux. The DS4300 is a computer in its own right and it sits in your SAN. You carve up the disks into Logical Units (LUNs) and I imagine you make those LUNs available to your Linux server based on the World Wide Name of the fibre adapter (the WWN is equivalent to a Mac address on a NIC).
Your server should then just see the disks as normal disks, so if you reboot the server it should see sda, sdb etc. or whatever disks there are (check dmesg).
So far so good. The extra step is to install the driver software that tells Linux the special things about these disks. For example, with SAN disks you often have two paths to the SAN and Linux sees two separate disks which are really, under the covers, the same disk. IBM will have some software that figures out the two are really the same disk and handles failover (so if one path fails, you carry on using the other path) and give you more commands to report on it all. For IBM's older SAN storage - the Enterprise Storage Server (ESS) that software was called SDD; I don't know what it is for the DS4300, but you should be able to download it from the IBM website.
|