Network programming.  Sockets, dial, UCX and Multinet.  Ask away.

11 responses total.

A co-irker of mine convinced me to start working on my CCNP, rather than
 renew my CCNA again. The material is not hard, it's all stuff I've
dealt  with before. Now if I could get the "around to it" to do the
damned  boring, simplistic labs. Server-builds and consultations for the
software  developers seem to sap all of my focus and get-up-and-go.
Anyone have some  spare motivation I can borrow?   -DTK

Learn Lisp.  All else will follow.  :-)

I do not consider counting close-parens to be the best use of my time. 
While I appreciate the idea of functional programming, I cannot get my 
head around it enough to generate functional codes. 

Cross, tag me off-line, I have an interesting turn of events to tell 
about. 

Counting parens?!  Heavens no; that's what paredit mode is for.

I'll shoot you an email.

Anyone screwed with IS-IS? (I know you carrier-network administrators 
have). I am interested in it because of SPB, but even my CCNP instructor
 (taking BSCI right now) has only minimal exposure to it. By only having
 two area types, does it make simpler to manage, or limiting? Does
moving  the area boundary off the router and onto the link make
significant  distinction? 

Is SPB as cool as it sounds? The speed and simplicity of a bridged 
network, but with the compartmentalization of a routed network sounds 
nifty. Do you find yourself growing to broad, flat, mushy networks? 
Getting rid of spanning-tree sounds fantastic, but what do you lose? 

Nothing is ever as cool as it sounds.  :-)

I've never used IS-IS directly, but have read about it in Radia
Perlman's book (which I highly recommend).  She describes the
circumstances of its creation and its relation to, e.g., OSPF and
other link-state protocols and how it grew out of work done for
DECnet and OSI, but later adopted for use on the Internet; it's
on par with OSPF.

SPB sounds cool; anything is better than the old spanning tree
protocol.

Reading up on IS-IS some more, it looks like you can have multiple
disjoint  backbones (level two areas), leading to stringy network
topology. What OSPF  does exceptionally well is impose a disciplined
approach to networking, in  which your topology has north-south
interconnect between east-west regions  enforced by protocol. 

Still, almost anything that is not spanning-tree would be better, as
long as  it avoids bridging loops. 

Just heard about Cisco's VIRL. WANT! When will it be out? It sounds like
 IOU/IOL, with a Nexus simulator and a BFR simulator. It would be nice
to  be able to model a real large-scale network, without investing
multiple  tens of thousands of dollars, or browning-out the whole
apartment  complex. 

re #8
I have a $450 setup for CCNA..5xrouter 3xswitch
It's the CCNAv3 200-125 exam lab with single homed eBGP and they've
added MPLS, MetroE, PPPoE and VPNs.
No Internet2 shenanigans though..c'est la vie
2811, 1841, 2950-24, etc

A nice application for vendor-neutral network testing and research is Mininet
(mininet.org). I've used it to test TCP multipath under various conditions.

NICE

Response not possible - You must register and login before posting.

- Backtalk version 1.3.30 - Copyright 1996-2006, Jan Wolter and Steve Weiss