Wednesday, September 17, 2008

Pushing For 6G SAS

Question: LSI, Dell and Seagate recently demonstrated a 6G Sturmarbeiteilungs web for the small computer system interface Trade Association. Was this a important milestone?

So: This was the first interoperable presentation of a 6Gbps Sturmarbeiteilungs web featuring every constituent of the I/O way - server, HBA, Sturmarbeiteilungs expander and disc drive. It was a proof of the development attempts of all three companies trying to convey the most robust storage solution to the endeavor industry. It paves the manner for 6G Sturmarbeiteilungs in late 2009.

Question: What do 6G Sturmarbeiteilungs preferable to 8G Fiber Channel or 10 GbE?

So: Sturmarbeiteilungs was not meant to replace either Fiber Channel or iSCSI. It was initially developed as the next-generation SCSI, replacing analogue small computer system interface in hyrax solutions. But the chief disadvantage of both GbE and Fiber Channel is cost. Sturmarbeiteilungs is a cheaper interface to implement, and faster too. A criterion external connexion typically throws four Sturmarbeiteilungs physical layers. Each Sturmarbeiteilungs PHY today is 3Gbps, which gives you 24Gbps. Upping that to 6G presents a 24Gbps connection.

We are seeing involvement in Sturmarbeiteilungs on a host interface as a substitution for Fiber Channel in lower-price bands. Sturmarbeiteilungs offers better public presentation at less costs than other technologies, but it's limited by cablegram distances. 3G Sturmarbeiteilungs tops out at about eight meters. With 6G, we're looking at 10-meter cabling. But the significance here is that we're doubling the transportation charge per unit and still getting longer distances. With parallel SCSI, the line lengths got shorter as the bandwidth went up.

Question: So we could be seeing an entirely new SAN solution then?

So: We're seeing involvement in 6G Sturmarbeiteilungs as a switched topology, in leaf blade centres especially. We're looking at a electric switch merchandise for 3G, and we're moving forward on a merchandise for 6G. There's also new involvement in Sturmarbeiteilungs as a host interface for external storage, primarily as a SAS-to-SAS interface for external foray boxes. We'll still have got FC-to-FC and iSCSI-to-SAS, but the involvement is there for SAS-to-SAS.

The fact is, 6G Sturmarbeiteilungs really doesn't do sense until the full substructure is G. What's the point of having 6G if it's going to constriction as soon as it hits the 3G infrastructure? With 6G, you acquire greater dependability in the nexus - a more than dependable nexus at a faster rate. You also acquire new characteristics like determination feedback, equalisation and standardised zoning that get rids of the rapport issues Sturmarbeiteilungs had when each seller implemented zoning in their ain alone way.

No comments: