Entry-level HP SAN array stoops to conquer small biz bods Once upon a time, HP had an MSA entry-level SAN array line-up with MSA 1000 (2gig FC) and MSA p2000 (8gig FC, 10gig iSCSI, 6 gig SAS) products. Now again it has two MSA array products: the MSA 1040 and 2040 dual-controller array products. Updated StoreEasy and StoreOnce product features were announced at HP's channel event in …
Sometimes quite a bit, sometimes not a lot.
A NAS is always attached to the network where the storage sits behind a controller that has two halves. The first half looks after the disk array (JBOD/raid or whatever). The second half is the network side that presents the disk array as an NFS/CIF type share. The clients (which might be a server) are never aware of the disk topology that supports the share.
A SAN can be network attached, fibre channel attached or a number of other technologies that connect the disk array to a host. This attachment is usually through some sort of fabric/network that abstract the array from the host. It is this network (FC, FCoE, iSCSI etc) together with the storage and host interconnect that makes up a SAN. Sometimes the host knows directly about the disk array, but usually it doesn't.
Some NAS boxes can also do iSCSI or Fibre channel which is where you get overlap. Sometimes they are configured to do both simultaneously.
I think that makes sense....Hope it helps.
@JustNiz
A NAS is, generally, a file server-in-a-box - you connect to it's resources in exactly the same way as you'd connect to a shared folder. If you are familiar with file shares on (e.g.) a Windows server, then a NAS is a dedicated appliance that does the same thing without the overhead of a full server.
As such, you access the storage on a NAS using a file sharing protocol - SMB/SAMBA/CIFS or NFS.
A SAN presents storage in the same way that local, 'direct attached' storage ('DAS') is presented and is therefore accessed via disk protocols, such as SCSI (modified as iSCSI) and fibre channel.
One way I explain the difference is by looking at which device manages the file table and read/write operations.
With a NAS, these operations are managed by the NAS, The client requests a file and then the NAS looks up its FAT and then accesses the relevant blocks. The NAS knows what has been requested (a file) and returns a file.
With a SAN, the client looks up its own FAT table and then directly requests the blocks from the SAN. The SAN does not know what the client wants - it just gets a request to read or write blocks and dutifully accepts or retrieves those blocks.
You might think that the NAS is a superior concept because all the client has to do is request a file whereas with a SAN, the client has to actually consult the file table and figure out which blocks it needs. The point is that that those operations have to happen regardless, the question is where.
The more clients you have accessing the storage on a device, the more efficient it will be to have the clients directly requesting the blocks.
You will often find devices that provide both types of functionality so referring to a specific piece of hardware as a SAN or NAS might be a bit ambiguous.
The difference is not at the hardware level, since often nowadays the same box can do both. The difference is at the protocol level:
SANs share data at the block storage level. The data will be presented to the client device as a chunk of storage, and often managed similarly to local disks.
NAS's share data at the file protocol level (eg. SMB/CIFS, NFS), and the data will be accessed as a network share of some description.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
