Hi, applogies if this has been thrashed to death before, but is there any facility to employ 2 DSS units (eg. using Automatic Failover) that share the same backend SAS (or even FC) storage (eg. as provided by an external SAS/FC Raid unit), and ideally protecting both iSCSI and NAS services?
With DSS, we get iSCSI with "automatic" redundancy (via Automatic Failover), but NAS with "manual" redundancy, but this requires disk storage to be fully duplicated (and then kept in sync via volume replication), yes? I understand this has advantages in some instances, but not all.
Eg. for NAS redundancy using Microsoft gear, this could be 2 x Windows 2008 hosts in fail-over cluster mode, sharing access to a SAS Raid unit (supporting persistent reservations). This would give file-sharing with redundancy, but not iSCSI redundancy (not supported by 2008 in fail-over cluster mode?).
Last redundant option I can think of using DSS that doesn't duplicate everything is: a dual controller SAS Raid unit, double the disks in 2 groups so each DSS only accesses 1 set, & so the volume replication traffic would use diff controllers/disks.
I agree that having 2 SAS Raid controllers accessing the same disks is a scarey thought, but we would be using 2 SAS HBAs accessing the same (external SAS) LUN, which sounds like something EVMS (after a quick read up on it) could do, if it was still current.
So the final verdict is that...
- for iSCSI automatic failover, we must duplicate all storage yes?
- for NAS services though, with the use of async volume replication, we could use fewer bigger & slower SATA disks, if faster smaller SAS drives were on the "primary" side, yes?