Interesting links (2011-05-29)

The article of the week is indubitably (how’s that for a Scrabble word?) RFC 6250, describing some of the stupidities (politely called misconceptions) TCP/IP stack had to deal with.

Other great content I stumbled upon during this week (in random order):

Data center

Christoph Jaggi published a new revision of his excellent Ethernet Encryptor Market Overview documents: introduction, point-to-point and multipoint products. All you ever wanted to know about encryption (and much more).

Stephen Foskett had a great FCoE versus iSCSI presentation @ Interop. The same content is now available on his blog.

John Herbert wrote a lengthy In-depth description of TRILL. Definitely worth reading.

The multihop FCoE saga continues with another great post by J Michel Metz.

Greg (@etherealmind) Ferro discovered true Golden Eggs: in-depth comparisons of x86 server platforms.

Cloudy services

I’m not the only one allergic to marketing nonsense. Five reasons your product isn’t a cloud is hilarious.

FathomDB decided to build a truly resilient relational database. Some people are learning from the AWS outage and accepting the harsh reality: IaaS clouds cannot be made fully reliable (or at least not at reasonable cost).

Evernote described their SaaS architecture. Read & learn (also applies to apps, server and storage engineers).


Tore Anderson makes a good case for IPv6-only servers and stateless NAT64 in data center (slides 27+).

RFC 6180 describes four typical IPv6 deployment scenarios and use of transition mechanisms. Must read #1.

RFC 6127 focuses on various forms of NAT. Must read #2.

None of the above

John McManus joins a long list of people telling vendors that we need emulators to test our designs and configurations, not in the cloud. Not sure many vendors are listening.

Five lessons learned in remote work. If you ever get involved in work-from-home discussions, read this first.

Daniel Dib describes a fantastic little known functionality: RMON can trigger alarms based on values of any SNMP variable. Yap Chin Hoong describes how you can use it to monitor CPU overload in NX-OS.

Add comment