New EVA generation being announced?

[Johan Guldmyr]

wohoo! :) Thanks Hari!

[Johan Guldmyr]

elreg reveals some more :)

http://www.theregister.co.uk/2011/05/05/hp_p6000_data/

[Matt M Burgess]

As for the EVA v 3PAR discussion, both are great and both have unique benefits over other. Assuming the new EVA has SAS disks, iSCSI ports, online LUN migration and thin provisioning, the key differentiators are:

3PAR benefits over EVA:

- Synchronous replication over very long distances;
- Better VMware integration;
- A standalone tool for App integration (you don't need Data Protector;
- the abiity to sustain a controller failure without going to write-through cache mode (if you have four controllers);
- online migration of LUNs off other storage;


EVA benefits over 3PAR:

- Using a CLI is a choice rather than a necessity (friendly = good for some people);
- EVA cluster can use Network RAID;
- EVA controllers can be SVSP controllers;
- Blade matrix automated provisioning (though this isn't far away with 3PAR);
- Insight Recovery support;
- It's cheaper


As always, it's horses for courses. IMHO neither of these products are "better" than the other, though as of today the 3PAR is vertainly more scalable.

the percieved "controller bottlenck" with an EVA is not a problem today because the EVA in engineered for the maximum spindle count to be the performance limitting factor.

The real "problem" with this model is that in 5-10 years when every array has SSDs in it, a controller-limited array is very limiting indeed. The EVA and that design have a place in the market as long as SAS disks exist and as long as there are customers whose performance requirements are less than the controller limit.