Cisco Blogs


Cisco Blog > Cisco Interaction Network

Data Center Interconnect Show Notes

June 17, 2010
at 12:00 pm PST

This story, Data Center Interconnect Made Simple, is all about connecting multiple data centers together to highlight the virtualized data center model with application and server mobility. This mobility is driven by cloud computing business models, disaster avoidance as well as data center migration/consolidation/planned maintenance. Data Center Interconnectivity helps move virtual machines across pods or physical data center locations with zero business and application downtime. To accomplish this we are focused today on three areas where new innovation is expanding options heretofore not available.

- Layer 2 Extensions. How we can extend layer 2 across vast geographic boundaries and connect remote data centers where the applications operate normally because they see the connectivity they expect.  Historically, this meant STP over long distance and it was a nightmare. This changes with Cisco’s OTV or Overlay Transport Virtualization.  See this animation we created to get an idea of how this works. Watch the whole show to be schooled by our own Ben Basler.

 

- Layer 3 Extensions: They are alive and well. Incredible improvements in Advanced VPLS are very much worth considering in the right environment.  Amit Singh joing Jimmy Ray in the lab to whiteboard the options here.

- SAN Extensions: are needed to provide virtualized application, access to storage without violating application SLAs (I/O time, latency) regardless of where the application and Virtual machine reside before and after a move.  We can’t leave that out of course, Jeff Thomas from EMC joins us to illustrate the need and the ability to deliver cache coherency between multiple storage arrays in these situations.

- Routing Optimization: The impact of these technologies in how we design and implement our networks can be critical. Max Ardica has been working on design guides so we can get on top of the question and the potential pitfalls.

 

Tough job…but someone has to put all of the above together…so what about those notes?

The connection of at least two data centers is not new.  The ability to go beyond the justification of redundancy and now look at the workload benefits is one of our promising new opportunities.

Keys to the Show

  1. Opportunity -- Does the opportunity here make sense to you?
  2. Grounded in YOUR Reality -- Is YOUR existing situation taken into consideration?  Do you have existing investments that need to be reflected here for something like this to be succesful?
  3. Layer 2 for the WAN- look everything for layer 3 but without the headache.
  4. Multi-Site Queasy -- Have you recently begun to digest the idea of application compute mobility within a single data center and the thought of doing this on a multi-location basis start to make you uneasy? 
  5. Storage Mobility -- Do you feel hamstrung by your existing storage strategy such that none-of -this-will-make sense until that is addressed.

The best ‘one page’ to go to for more info on all of this when you are ready?  www.cisco.com/go/dci


The Challenges of Workload Mobility....with our new host Omar Sultan!

 

Omar Sultan Single

Omar Sultan identifies both the opportunity and the challenges that must be overcome.

 


OTV, Overlay Transport Virtualization with Ben Basler

Layer 2 adjacency may be needed for applications to function properly but the prospect of extending this across the LAN would historically be considered lunacy.  Ben Basler lays out the Cisco OTV solution.

OTV (Overlay Transport Virtualization) can be described as a MAC in IP technology that supports Layer 2 and extension over any transport. So I don’t care whether the core is IP, whether the core is MPLS as long as the core can actually deliver an IP packet from point A to point B, that’s fine for OTV. OTV actually is based on two technology pillars, dynamic encapsulation and protocol learning. And we will see what the benefits are of these two pillars like no pseudowire maintenance, optimal multicast replication, built-in loop prevention and so on.”

Ben Basler on OTV

Overlay Transport Virtualization for VMotion

OTV Dynamic Encapsulation

OTV Protocol Learning

OTV Demonstration

OTV Demo2

 


A-VPLS, Advanced Virtual Private LAN with Amit Singh

 

Advancements in layer 3 connectivity continue to be addressed as Amit Singh reveals the latest in VSS, load balancing and simple configuration.

Feng Meng covers the benefits of this strategy quite well in this Data Center Blog entry.

Very nice ‘First Look’ from InfoWorld’

 

 

 


SAN Interconnect with Jeff Thomas from EMC

 

Special guest Jeff Thomas from EMC reveals VPLEX recently announced at EMC world that bring virtualized storage into reality.

More on VPLEX

Rick Sherer over at VMWareTips.com shares his thoughts.

Chad Sakac, whom we are still trying to match schedules with always explains these things well with his entry at Virtual Geek with ‘Your Virtual Machine Teleporter is Ready’

He put together a nice demo, very similar to what Jeff covers in our show…but you may find helpful…and you should be following Chad’s blog anyway. 

 


Routing and Path Optimization with Max Ardica

All this talk about layer 2 can’t ignore the changes you need to consider for this affects your routing strategy…

Implementation and Design Guidance (.pdf)

 
Max Ardica in the Lab

Comments Are Closed