HP EVA and RSS representation

[aeronet]

Hi,

for better understanding of the EVA concept with the Redudant Storage Set I have drawn it (look at the attachment).

Can anybody tell me if the representation is correct.

Thanks

[IBaltay]

Hi,
this thread can help you to check your scheme pls:
http://forums11.itrc.hp.com/service/forums/questionanswer.do?admit=109447626+1218795782995+28353475&threadId=886230

[Uwe Zessin]

No, I don't think this is correct. Mirrored data is stored on adjacent disks. So "A1" is stored on RSS members 01:00 and 01:01, "A2" on 01:02 and 01:03 and so on.

The data structures are a bit more complicated, so A2 could be on 01:00 and 01:01, too. And a VRAID-5 stripe need not be on 5 consecutive disk drives - it could be stored on any 5 different disk drives within an RSS.

In the legend, it should be "RSS ID/INDEX"

[aeronet]

Ok,

in the attachment a corrected version.

> The data structures are a bit more
> complicated, so A2 could be on 01:00 and
> 01:01, too.
> And a VRAID-5 stripe need not be
> on 5 consecutive disk drives - it could be
> stored on any 5 different disk drives within
> an RSS.

But _could_ be theoreticly also in the order I have drawn in my scheme.

[Uwe Zessin]

> But _could_ be theoreticly

Agree. Nobody can really tell unless (s)he looks into the current mapping tables (which are not accessible to the end-user) and even the mapping can change over time. But for illustrations I draw it the same way, by the way ;-)