[Lustre-discuss] iSCSI Integration Solutions

[Brian J. Murrell]

On Tue, 2009-09-29 at 10:24 -0400, Pete Eby wrote:
> 
> Your are correct in that I have configured a
> luster installation and want to re-export the file system from luster to
> other clients via iSCSI  - and I can't quite seem to grasp how this can
> be done.

That's because it can't.  :-)  Lustre is a filesystem.

> Yea, I don't expect Luster itself to present iSCSI, but rather how take
> advantage of an presented luster FS and re-export as iSCSI.

Well, if you can come with any other way of providing normal filesystem
namespace as an iscsi device, you might have a chance.  But until you
can point to something in say /export/ and make it an iscsi target, you
won't have any hope of lustre doing this for you.  But TBH, I think you
are barking up the wrong tree.

> My goal (like others before me I see from Google) is to use the benefits
> of Lustre as a clustered, scalable storage solution and re-export it
> through iSCSI (which will eventually be used by VMs).

Why not just make the clients Lustre clients instead of iscsi consumers?
What are these iscsi consumers (VMs did you say?) going to do with the
block device once they get it?  Put a filesystem on it?  What OS will
these VMs be running?

> It seems like this is possible,

I doubt it.

> but I am missing how to bridge between
> say a Luster client, connecting to an exported Luster FS and re-offering
>  that file systems through iSCSI.

As I said.  Unless you can make this work with any old filesystem
namespace on Linux, you won't get it working with Lustre either.

b.

-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 197 bytes
Desc: This is a digitally signed message part
URL: <http://lists.lustre.org/pipermail/lustre-discuss-lustre.org/attachments/20090929/c2317709/attachment.pgp>