| Q1. |
What is the DC SAN Backbone Director? |
| A1. |
The DC SAN Backbone Director is a new class of network infrastructure that resides at the core of the Data Center Fabric. Designed to meet the growing connectivity, virtualization, and cost-efficiency needs of enterprise data centers, it delivers breakthrough performance, a highly robust services platform, seamless interoperability across fabrics and networking protocols, and future-built extensibility. |
|
| Q2. |
What is unique about HP’s DC SAN Backbone Director? |
| A2. |
DC SAN Backbone Director is a distinctly new class of network infrastructure that is complementary to HP’s B-series director and switch product family. Designed for the largest enterprise data centers as well as vast virtual server environments, the DC SAN Backbone Director delivers 5x the performance of director-class devices and broad capabilities beyond high-speed, six-nine reliable SANs.
The DC SAN Backbone Director delivers a robust, intelligent platform for plug-in, network-based services, a high-performance, high-capacity connection point between heterogeneous SAN (FC, FICON) fabrics, and application-centric, and policy-based QoS for dynamic virtual server and virtual storage environments. DC SAN Director is also future-built to enable the merging of multiple types of networks (storage, client/server, virtual server cluster, extension, file, FICON) with high bandwidth/low-latency, lossless connectivity.
DC SAN Backbone unparalleled performance, scalability, interoperability, and intelligence help large enterprise data centers to improve server and storage utilization, reduce costs, and achieve non-disruptive growth.
|
|
| Q3. |
Why do customers need a SAN Backbone Director now vs a SAN Director? |
| A3. |
To address performance, data security, virtual server, and fabric connectivity needs today, while enabling non-disruptive adoption of emerging network technology with investment protection tomorrow. The Data Center Backbone is an investment for today and for the future. |
|
| Q4. |
What type of Multi-protocol support does the DC SAN Backbone offer? |
| A4. |
There are several network protocols leveraged in the data center and connected SAN, FICON, HPC, and Ethernet islands. DCX delivers 1 – 10 Gbit/sec Fibre Channel, FCIP, iSCSI ready, today and is 10GE (along with FCoE) ready as well as able to accommodate future HPC protocols such as CEE. |
|
| Q5. |
Does FCoE replace Fibre Channel? |
| A5. |
No. FCoE is Fibre Channel run on a new Data Center protocol based on 10GE and leverages the copper cabling plant of traditional Ethernet. FCoE will initially be used to connect server islands into the traditional SAN. FCoE is not expected to replace existing fiber channel or be used for future storage area networks supporting tier 1 applications for several years. |
|
| Q6. |
How does the DC SAN Backbone help virtual environments? |
| A6. |
Besides providing unprecedented bandwidth to the growing number of applications being attached to shared storage, DC SAN Backbone provides Adaptive Networking features that ensure service levels are maintained as applications move in a virtual infrastructure. Policy based services are automatically adjusted and reapplied as an application moves from one physical asset to another. |
|
| Q7. |
How does the DC SAN Backbone Director provide 6.5 Terabits of Aggregate Bandwidth? |
| A7. |
384 ports x 8.5Gbit/sec (line speed) x 2 (full duplex) = 6.528Tbit/sec. DC SAN backbone Director’s advanced Backbone combined with Brocade’s local-switching capabilities effectively eliminate over subscription. (Aggregate bandwidth for the Cisco 9513 is only 1.1 Tbit/sec).
|
|
| Q8. |
What is the Inter Chassis Link bundle, and what do customers gain by using it? |
| A8. |
Customers gain 128 additional external ports by using ICLs connecting 2 DC SAN Backbone Directors to each other eliminating ISLs.
Each ICL has the following characteristics:
- 2 meters long proprietary cable
- Connection is through the core switching, each blade has receptacles for 2 ICLs
- All connectivity patterns for the ICLs are equivalent for performance and availability
Each ICL connection is the equivalent of 16 fixed speed 8Gbit/sec E_PORTs.
ICLs yield a 512Gbit/sec connection between 2 DC SAN Backbone chassis’. This is a full-duplex connection so the 512Gbit/sec can simultaneously flow in each direction for an aggregate data movement of 1Tbit/sec.
ISL Trunking is automatically enabled across the 16 x 8Gbit/sec connections within each ICL resulting in a 128Gbit/sec, frame-balanced pipe. HP DPS operates with ICLs so exchanges are balanced across the 4 ICLs using a hash of the SID, DID, RxPort, and OxID.
ICLs can work in conjunction with conventional ISL using optical cables and ports on the DC SAN Backbone Director port blades. ISLs can be used to add bandwidth to the ICLs or to connect other switches to a DC SAN Backbone Director.
|
|
| Q9. |
Is the DC SAN Backbone Director blocking? |
| A9. |
No. The DC SAN Backbone Director architecture does not drop frames under any circumstance, so it does not “block”. It is possible to congest a DC SAN Backbone chassis if more than 256 ports need to simultaneously switch at 8Gbit/sec or a single blade has more than 256Gbit/sec of Backbone bandwidth requested. |
|
| Q10. |
The DC SAN Backbplane requires B-series SFPs, Why? |
| A10. |
Given the nature that SPF tolerance decreases as frequency increases, HP has standardized on specific specifications for it’s B-series SFPs. This ensures reliable connectivity and reduces support calls. |
|
