Welcome to Infotrek, the newest podcast in the Packet Pushers Community Channel!
Infrastructure engineers think about and deal with redundancy every day. We implement it on everything we touch without considering its inefficiencies, complexity, and whether we should even be using it in a particular situation.
In this episode, after talking over some news items, John, Derek, and Mike dive into what redundancy really is, why we use it, when to not use it, and implementation challenges.
John W Kerns is a Hybrid Engineer focused on network and data center (systems, virtualization, etc.) technologies.
Derek Pocoroba, CCIE 18559, is the Director of Enterprise Networking and focuses on network architecture and design for customers.
Mike Aossey is the Director of Practice Management and works day to day overseeing SMEs and architecting VoIP solutions.
John, Derek, and Mike are all colleagues at a VAR based in Southern California.
Oracle buys Ravello – Fortune
Cisco flexes its HyperFlex hyperconverged muscles – The Register
Show Topic: Redundancy
-What is infrastructure redundancy and why is it important?
-When is redundancy necessary?
-When is it appropriate to not use it?
-How is it usually implemented?
-What are typical implementation challenges?
- N+1 is very inefficient when N=1 (you can only use 50% capacity)
- As N becomes larger, a N+1 redundancy model becomes more efficient
-Is nested redundancy necessary?
- For example, dual supervisors in a pair of chassis in VSS
- Dual power supplies in a failover device pair
-How can redundancy be tested after it is implemented?
- What are typical problems found during testing?