DSS and Solid State Disks (SSD)

G'day All,
A recent post got me thinking about the way DSS works and SSD.

We have recently deployed a DSS with 6x 1TB ES.2's in RAID 10 and 2x 32GB in RAID1 to support an ESX 3i configuration.

So far the performance has been very good and the client is impressed. (I will post proper specs once we have it finished off.)

But I started to think about the differences with SSD and in particular things like wear leveling and the relative small size of SSD vs SAS/SATA.

So my question to everyone is:
What optimisations do you think we should be applying for the SSD volume?
I did a search for SSD in the KB, but nothing showed up.
Should a feature request for DSS be to be able to choose a volume "type" to provide defaults that are suitable for the underlying technology?
For example, should the swap file be eliminated (or placed on a different volume somehow)?
Are there particular applications that SSD is not suitable for?

Given that SSDs are rapidly increasing in size and dropping in price, I do not think it long before we will be deciding between SSD vs SATA, not SAS vs SATA.

Rgds Ben.

One other thought along these lines is "tiered" storage as well.
ie shifting older data onto slower media when it is unused/older. I realise that this would probably be hard(impossible) to do for iSCSI but for NFS/SMB it may be possible?
Since XFS supports DMAPI, is this possble at all?

Dear Beng,

We don't have much information on the SSD drives, as it is not tested yet at our labs. It's going to be added on the roadmap, but currently the roadmap is full for the next 2 or 3 quarters. But a request going to be send to added and searched by our developers team.

Regards,
SJ