This blog is to help to identify new Cisco Nexus Product family. Specifically the Cisco Nexus 7000 switch chassis and components. I will also talk about Cisco Nexus 5000 switch, and the Cisco Nexus 2000 Fabric Extender and Nexus 1000V .
Lets start at the bottom of the tree.
Nexus 1000V: In this new world of virtualization we need to be able to connect to many virtual machines on the network without compromising the throughput and performance on the network layer. Nexus 1000V is a Virtual Switch connects these virtual machines. It’s easily integrates with Vmware environment. Nexus 1000V runs same NX-OS operating system same as other Nexus switches. Though it’s virtual switch, it still provides the rich feature sets in a virtual environment.
Once it’s installed and operational, the control of the switch is back in the hand of the network administrator, instead of relying upon server admin . Network administrator can configure and manage this switch same way as he/She would do a normal switch.
So, how does this forward traffic ?
Once installed, it replaces the VMware’s virtual switch component. The data plane is called VEM or Virtual Ethernet Module. VEM installs directly to ESXi Hypervisor. For control plane and management plane functionality it uses something called VSM or Virtual Supervisor Module. This VSM can be hosted on ESXi as a Virtual Machine or on an appliance called Nexus 1010.
VSM does not get involved in data transfer or connectivity, that is VEM’s job.
Getting complicated ? Just think about a Modular switch 6500. When we do show module , we see the supervisor module and all other modules installed on them. Similarly here when we do show module on VSM, we see Virtual Supervisor Modules and Virtual Ethernet Module. We can have multiple Virtual Supervisor Modules to be as Hot Standby for failover scenario. Failing of one Supervisor does not affect the VEM or Data Transfer through VEM.
Think about this, Cisco 1000 V is an access switch connecting the servers..(just on a virtual environment )
Cisco Nexus 5000 Series :
This is cisco’s first offering of high density, low latency 10 Gig/sec Ethernet switch. This are Layer 2 switches.
It has Redundant Hot swapable Power supplies, Redundant hot swapable fans. Cooling is front to back . This is good to deploy in a hot isle/cold isle data center. N5K supports both FCoE and Native Fiber channel. Which is good if you have older fiber channel environment but would also like to have FCoE functionality in today’s environment.
There are two Models in N5K family.. Nexus 5010 and Nexus 5020 .
Nexus 5020 can support upto 56 Ports. Out of these 40 ports are fixed. And it has two expansion modules. You can install modules according to your need.
Nexus 5010 has exactly half the ports. It has 20 fixed ports with 1 expansion bay.
We can also use Nexus 2000 series Fabric extender with N5K switches. Fabric extender works as a extended Module of the Nexus 5K or Nexus 7K switches.
There are two new models introduced recently in Nexus 5000 series. They are Nexus 5548(with 36 10G ports and 1 expansion module) and Nexus 5596(with 48 10G ports and 3 expansion Module)
So based on the requirement and port density we can choose either of these switches.
Nexus 5500 series adds exciting features which earlier Nexus 5010 and 5020 did not have. 5548 and 5596 series have Layer 3 routing capabilities where 5010 and 5020 were Layer 2 only switches. This also support Cisco Fabric Path technology which was only available on Nexus 7K before.
Here’s the chart showing the improvements of Nexus 5500 over nexus 5000 series switches. Take note of the Throughput, VLAN config and Port to Port Latency between the models..
Cisco Nexus 7000 Series:
You can call it the next big thing, the Evolution . This was designed ground up to support modern day networks, with high throughput, scalable, modular switching technology. Without a doubt it was built for datacenters.
Nexus 7000 has 3 models..
Nexus 7009, 7010 and 7018 .
Have a look at the below Chart to see the difference between them..
In depth look at 7009 Chassis..
The modules are placed horizontally, so the airflow must be side to side. Cable management are built in both sides to offer cable run from both sides. The front door can be locked to prevent any accidental cable movement. 7009 isn’t as deep as 7010 or 7018.
Next have a look at Nexus 7010 Chassis..
This was built for power and cooling optimization and resiliency. Every component in Nexus 7010 is redundant. I/O modules are mounted vertically, this allows Front to Back air flow. It has cable management option at the top of the rack with is really cool. It also has the option to lock the doors at the front to prevent accidental cable disruption.
Finally.. the big boy.. Nexus 7018 Chassis..
It’s an 18 slot chassis, including 2 supervisor module. It leaves us with 16 slots for I/O modules. It also has Side to Side Airflow and integrated cable management. Again , everything are redundant. From Fabric modules to Fans to Power supply.
The Fans rotations increases or decreases according to need, which saves electricity. Here’s how it looks ..
Woooh.. thats all about the chassis.. however we aren’t done yet !!. We need to know about the Line modules goes into these chassis right ? You can’t just buy the chassis not have any line modules ! That’d be like buying a box of chocolate without the chocolates in it !.. not a good example but you get the point..
Nexus 7000 Line modules..
First up .. Supervisor engine..
Nexus 7000 Supervisor engine(Sup 1)..
This is a true supervisor engine which controls management plane and control plane only. No switching is performed on the Supervisor engine. This is nothing but a management engine which is always on , N7K to be fully redundant , we need to have 2 supervisor engines.
So..what’s on this supervisor engine or how does it look.. here it is..
Next up is the Crossbar Fabric Modules..
Fabric modules are the component which provided high throughput to the I/O slots, ie slots that connects the cables. Those IO slots connects to the Fabric Modules on the backplane , which ensures the High Throughput.
Remember those 7010 and 7018 switches ? at the back it had 5 slots for Fabric modules. These are the modules I am talking about here.
So far we have seen both sup module and fabric module, now lets look at the I/O modules, which I call Port modules. Just so that I remember, these are the modules with the Ports on them and takes the data.
First up is 8 port 10Gig module. It’s called M1 module. It gives us 80 gig full duplex fiber connectivity. If we have other switches to connect or up-links, you can use this type of modules.
Second is 32-port 10Gig Module. It is also M1 module. It gives us higher port density for 10 Gig connectivity with Over subscription. Here’s how it looks..
Next up is Nexus 7000 48 port 1Gig I/O module..
It has 48 1 gig ports with either SPF or RJ45 . It looks like same as other Modules we have seen on 6500 switches.
Last but not Lease.. Nexus 7000 F1 series I/O module.. lol F1. I call it Formula One module. This module is used for High performance low latency throughput. A mixture of SFP and RJ45 ports is used. It also can be dual speed, ie 1Gig or 10 Gig .
F1 modules do not support Layer 3 functionality !! We will need to use M1 cards for Layer 3 !
That’s all about the Modules ..incase you are forgetting.. we have Nexus 2000 models , which are used as Fabric extenders. Now think about it, you might have 1000s of servers in a datacenter which needs to be connected to the network. We aren’t going to install N7Ks everywhere, we will install 1 at the end of the Rac, and use Nexus 2K on top of each rack to connect the servers. This 2Ks are nothing but an external Modules to the 7Ks. That’s the beauty !..
And this is how it will look..
That’s all for today !!! Hopefully in future I will talk about the configuration and troubleshooting of Cisco Nexus devices.