Posts Tagged 'Data Center'

Virtualized Data Center Webcasts

Published February 19, 2010 Data Center Leave a Comment
Tags: ,

Last month, Cisco, VMWare, and intel put together a virtual conference on data center virtualization. The recordings of the 13 webcasts presented are now available online. Topics covered were:

  • Virtualized Data Center Roundtable
  • Emerging Trends in Enterprise Cloud Infrastructure
  • Virtualized SAP Applications: From The Datacenter To The Cloud
  • Enterprise Cloud Solutions – Focus on Flexibility
  • Transformational Business Value of Data Center Virtualization
  • Unified Computing System Innovations
  • Case Study: Terremark’s InfiniCenter Virtualization Story
  • Data Center 3.0: Unifying the Virtualized DC Infrastructure
  • Nexus 1000V
  • Cloud Computing: Will a Growth Story Unfold in 2010?
  • SAP Applications in Virtual Environments
  • HyperCED the new Data Center as a Service
  • Server Virtualization: Maturity Brings Complexity

Data Center Ethernet and the Nexus Switches

Published December 21, 2009 Data Center , Dennis Hartmann Leave a Comment
Tags: , ,

Small Computer System Interface (SCSI) hard drives have been the preferred method of hard drive storage for servers in enterprise networks for a long time. The Serial attached SCSI (SAS) standard is an electrical standard in which SCSI hard drives are chained together and identified by a SCSI ID. SCSI is an electrical standard which only allows devices to be connected up to 10 meters (32.8 feet) from the server.

Fibre Channel (FC) is an optical standard that allows SCSI hard drives to be connected at much further distance depending on the type of fibre optical cable and laser used to drive the signal. Small form factor pluggable (SFP) optical transceivers are used in director switches, storage arrays, and director switches to create a storage area network (SAN). SAN is sent over storage switches, while LAN data traffic is sent over data switches.

Fibre Channel over Ethernet (FCoE) is a technology that is aimed to replace the director switches of the SAN infrastructure by allowing servers to transfer both data and storage traffic over 10 Gigabit Ethernet interfaces using server network interface cards (NIC) that have both fibre channel and Ethernet chipsets.

Fibre Channel frames have a maximum payload size of 2112 bytes, but the FCoE encapsulation process results in a 2179 byte frame. Jumbo frames must be enabled on Cisco Nexus switches to accommodate the overhead of the Ethernet and FCoE header and trailers. Ethernet jumbo frames can accommodate frame sizes of up to 9,216 bytes (9K). The transmission of fibre channel traffic over a data center infrastructure represents some challenges that Ethernet was not originally designed around. Fibre channel does not tolerate frame loss. The Ethernet infrastructure that will transmit FCoE traffic must be designed to be lossless, resulting in zero frame loss.

Cisco Nexus switches are designed around data center bridging’s low latency and low loss requirements.  Cisco’s priority flow control (PFC) is an important technology that is based on the IEEE 802.3x link level pause frame, but the technology can now be implemented at the IEEE 802.1p class of service (CoS) level.  IEEE 802.3x pause frames are used to flow control Ethernet devices that are aggressive sending data when there is congestion on an uplink. IEEE 802.3x flow control operates at the link level causing all traffic to be delayed when this condition arises. The PFC implementation allows Cisco to flow control individual class of service (CoS) values similar to the IEEE 802.1Qbb working group. Since PFC handles congestion at the CoS level, FCoE traffic can be mapped to a CoS value that does not receive a PAUSE frame while the lower priority traffic receives a pause frame. PFC will help guarantee a lossless FCoE environment.

Cisco Nexus switches support a pre-standard version of backward congestion notification (BCN) draft IEEE 802.1Qau standard. BCN allows the Cisco Nexus switches to communicate congestion to end stations and server where PFC only allows congestion throttling at the CoS level on a link by link basis.

Data center bridging exchange (DCBX) protocol allows Cisco Nexus switches to dynamically discover the data center capabilities of other Nexus switches by leveraging negotiation parameters of the Link Layer Discovery Protocol (LLDP).

The Nexus data center switches provide a migration path to fibre channel over Ethernet at a competitive cost, while running at a lower power footprint than previous switching solutions.

Author: Dennis Hartmann

Data Center Ethernet

Published December 14, 2009 Data Center , Dennis Hartmann 1 Comment
Tags: ,

Data center environments have two infrastructures currently. Servers in the data center are connected to the local area network (LAN) and storage area network (SAN) infrastructures. Fibre Channel over Ethernet (FCoE) provides a mechanism to transfer both fibre channel and network traffic over the same unified Ethernet fabric.

Servers connect to the LAN infrastructure through at least two Gigabit Ethernet (1Gbps) network interface cards (NIC) teamed together to provide redundancy and higher throughputs. Most high performance computing servers have six to ten Gigabit Ethernet interfaces to provide adequate throughput to the server. The Cisco 6500 series switch has been the most popular LAN switch in data center environments for nearly a decade. 10 Gigabit Ethernet (10GE) is a technology that has been used over the past eight years to provide higher capacity uplinks between switches.

High performance servers utilize SCSI hard drives that have a distance limitation of about 10 meters (32.8 feet). Fibre channel was designed to overcome the distance limitations of SCSI, while utilizing SCSI drives for storage by using SCSI as an upper layer protocol. Fibre channel (FC) host bus adapters (HBA) provide connectivity to the SAN through director class switches sold by EMC and other storage vendors. Fibre Channel HBAs operate at speeds of 1, 2, 4, and 8 Gbps, but 8Gbps FC HBAs are not widely adopted at the time of this writing.

FCoE converges the LAN NIC and SAN HBA functionality into one 10GE NIC card that provides an HBA and 10GE ASIC (application specific integrated circuit). Emulex and QLogic are the pioneers of this technology at the time of this post. The 10GE NICs are available as a single or dual port solution. The 10GE interfaces connect to FCoE interfaces on the Cisco Nexus 5000 switches, which will provide connectivity to both the LAN and the SAN, eliminating the need for FC director switches and HBAs saving real estate, power, and cost in the data center. Servers utilizing the PCI express 1.1 (8x or 16x) standard will support wire rate 10GE speeds.

Converging storage and network traffic onto the same infrastructure provides a unified I/O (input/output) fabric. The Nexus 5000 series switch provides 20 (5010) or 40 (5020) 10GE interfaces for data center network connectivity with the option of one (5010) or two (5020) expansion slots that allow the following connectivity options allowing enterprises to migrate to FCoE:

  • 8 port FC card (1/2/4 Gbps supported)
  • 4 port FC card / 4 port 10GE FCoE
  • 6 port 10GE FCoE

Author: Dennis Hartmann

References

Cisco Nexus 7000 and Virtual Device Contexts

Published October 30, 2009 Data Center , Dennis Hartmann Leave a Comment
Tags: ,

The Cisco Nexus 7000 data center switch provides layer 3 capabilities in the Nexus data center product lineup using the Cisco NX-OS operating system. The Nexus data center switch is the first Cisco switch to provide virtual device context (VDC) capabilities that allow the data center switch to be logically segmented into up to four different switches (device contexts).

When the Nexus 7000 switch is first booted, all interfaces of all line cards in the modular switch platform are placed in the default VDC (VDC 1). Up to four VDCs can be created, including the default VDC, but an interface can only belong to one VDC. Interfaces from different line cards can be placed in different VDCs. Once an interface is placed into a VDC, the interface can no longer be managed in the default VDC context.

Each VDC is configured by name, but a VDC identifier is assigned automatically in the order in which the VDCs were configured. The following configuration will create a VDC called Highpoint and assign all 32 ports of line card 18 in the Nexus 7018 10GE line card:

  • N7000#config t
  • N7000(config)#vdc Highpoint
  • N7000(config-vdc)#allocate interface Ethernet 18/1 – 32

Ethernet ports 18/1 – 32 have been assigned to the Highpoint VDC and will not show up if the show interface brief command while in the default VDC context.  The show vdc command will display all VDC names, VDC IDs, and mgmt 0 interface MAC address. The Mgmt 0 interface is an Ethernet-based interface on the supervisor module of the platform. The show vdc membership command will display all VDCs and the interfaces to which they are applied.

To switch the configuration mode from the default VDC to the Highpoint VDC context, use the following commands:

  • N7000(config-vdc)#end
  • N7000#switchto vdc Highpoint

It is recommended to change the hostname of each VDC to change the configuration prompt to indicate which VDC you are configuring. NX-OS uses the switchname command instead of the standard Cisco IOS hostname command. To switch back to the default VDC context, use the switchback or exit commands. Each VDC is managed as if it were a separate switch.

Guest Author: Dennis Hartmann

Editor’s Note: You can read more from Dennis on the Global Knowledge Unified Communications blog.

Nexus Data Center Switches (NX-OS)

Published August 3, 2009 Data Center , Dennis Hartmann 6 Comments
Tags:

The Cisco Nexus switches allow a very high level of bandwidth connectivity positioned as data center switches. The Nexus switches do not support the same features and flexibility as local area network (LAN) based switches, but provide the very high capacity and cost effectiveness needed in data center environments. Nexus switches run a new operating system called Cisco NX-OS and provide various Ethernet based interface speeds varying from 100Mbps to 10Gbps depending on the Nexus platform. The Cisco Nexus product line currently consists of the following product families:

  • Nexus 1000
  • Nexus 2000
  • Nexus 5000
  • Nexus 7000

All Nexus platforms run the Cisco NX-OS operating system that is very familiar to Cisco IOS, but offers much more flexibility and reliability than Cisco IOS. The Nexus 1000 supports many virtualization technologies that are in VMWare ESX platforms and VMWare’s newer vSphere technologies. The Nexus 2000 and 5000 platforms provide layer 2 connectivity, while the Nexus 7000 provides layer 3 connectivity.

This blog will focus on a deployment model leveraging the Nexus 5020 and Nexus 2148 devices. The Nexus 5000 platforms provide non-blocking backplane architectures with onboard capacity of twenty (5010) or forty (5020) 10 Gigabit Ethernet interfaces with one to two expansion slots respectively. The expansion modules have the ability to provide additional 10 Gigabit interfaces and fiber channel over Ethernet (FCoE) connectivity. The first 8 port (5010) or 16 (5020) ports of the 5000 platform can be configured to provide 1 Gbps connectivity using small form-factor pluggable (SFP) optics, while all the ports can provide 10 Gbps connectivity over SFP+ optical interfaces. The Nexus 5000 platforms support connectivity to the Nexus 2148 through a fabric extender (fex) configuration as of NX-OS 4.0(1a).

The Nexus 2148 platform does not have a local console port or any way to directly configure the device. The Nexus 2148 data center switch is managed through the Nexus 5000 configuration. The Nexus 2148 appears as a line card in the configuration of the Nexus 5000 after the fabric extender interface is properly configured and comes online. The Nexus 2148 downloads NX-OS software from the Nexus 5000. After the 2148 is online, 48 new configurable interfaces appear in the Nexus 5000 configuration.

The Nexus 2148 platform provides 48 Gigabit Ethernet interfaces and up to four fabric extender (fex) interfaces to the Nexus 5000 platforms. The Nexus 2148 supports up to four 10 Gigabit Ethernet (GE) fabric extender interfaces to the Nexus 5000 platform, but the Nexus 2148 can only uplink to one Nexus 5000 with NX-OS 4.0(1a). NX-OS 4.1(3) supports redundant uplink connections to two different Nexus 5000 switches providing interface and path redundancy.

IEEE 802.1d Spanning Tree Protocol (STP) is enabled on the Nexus 5000 by default and all redundant paths between the Nexus 2000 and 5000 are placed in the blocking state by STP. Two channel groups can be configured to each Nexus 5000 which will result in two forwarding (channel group 1) and two blocking interfaces (channel group 2) which still results in sub-optimal bandwidth utilization.

Virtual port channel (vPC) technology introduced with Nexus 5000 NX-OS 4.1(3) allows the 2148 platform to use all four interfaces uplinked to the Nexus 5000. vPC technology is similar to the virtual switching system (VSS) technology introduced in the Catalyst 6500 VSS supervisor module which is similar in nature to multi-chassis Etherchannel (MCEC). All the above technologies allow the uplink switch to appear as one virtual switch to the downstream devices. Loop avoidance is provided through a vpc link interface between the Nexus 5000 switches which requires a special configuration. Path recalculation is performed in under one second on the Nexus platforms in the event of a link or device failure.

References:
Data Center Switches

Cisco Nexus 5000 Series Switches

Special thanks to Guest Author Dennis Hartmann for this post. You can read more of Dennis’ work over on our Unified Communications blog.

Subscribe in a Reader

Subscribe by Email

Follow us on Twitter

Featured Courses

Archives

Cisco Live 2010