Even thought Mr Paton has pipped me at the post by working late at night I have to put something official on here from Oracle regarding the new models from Oracle - and he's probably still in his bed at this time
So you will see from the link below that we have announced some new models in the ZFS Storage Appliance range (new name for the family).
There will be 3 new models available from the middle of October 2010 starting with the new S7120 that starts at 12TB and scales out with the addition of the new SAS2 disk shelves to 120TB (using the new 2TB SAS drives). In addition the S7120 supports write flash capability using the F20 card which is also deployed to great effect in our standard servers and the Exadata device.
The next model is the S7320 which starts at 20TB and goes out to 192TB and supports both read and write flash giving true HSP capability. You can also cluster the S7320 for even greater resilience.
The final model that is being released is the S7420. This, like the S7320, starts at 20TB but with the addition of extra SAS2 disk shelves scales out to 1.1PB - which should hopefully be sufficient for you edugeekers :-)
These new ZFS Storage Appliances are all based on the latest Intel Nehelam chip and can support up to 1TB of DRAM (S7420) and use the SAS2 drives to deliver even greater throughput and performance.
If you need any further detail or information then please do not hesitate to contact me.
Thanks Phil, sounds very shiny stuff Looking forward to hearing how they get on in the education market
Last edited by john; 21st September 2010 at 08:40 AM.
Whoop, finally! Been waiting months for this... Talking with Andy @ Cutter about SGD in half an hour, however I feel the conversation may rapidly deviate to an S7120!
Phil - Would a 7120 with the new Nehalem, an F20 and 16GB RAM (or is 32GB the minimum now?) likely offer better performance than our current 7410 with Opterons, 16GB RAM, 2 Logs and 1 read? Just wondering which would be best for the primary unit in a mirror?
@Duke: Surely you want to get another 7410 so that you can cluster them
Zero downtime with clustered heads would be great, but a minimised downtime of the server room going up in flames and it falling over to a 7120 elsewhere on site is more important to us from a DR/business continuity point of view. Clustering only saves me if a head fails - if the disk tray dies, power goes out, or a plane crashes into the room I'm still screwed.
Just spoken to Andy at Cutter and these things may be silly cheap for enterprise storage!
Soulfish (21st September 2010)
Nope! - I don't think Phil's speaking to me any more after Andy at Cutter told me thing he shouldn't.
There are currently 1 users browsing this thread. (0 members and 1 guests)