Interesting links (2011-02-06)

Numerous interesting technical articles I found during the last week:

The loopback interface is always reachable ... or maybe it’s not, depending on your software release. In totally unrelated news, Gartner is telling your CIO how a multi-vendor network reduces the overall costs. Maybe it’s time to factor all the hidden costs in the calculation. BTW, while you’re reading Gartner reports and talking to your vendors, keep this list handy.

Joe Onisick wrote a fantastic Disaster Avoidance introductory article. If you need the basics, this is the first thing you should read. If you need more Data Center Interconnect (DCI) details, register for my DCI webinar.

Dave Alexander has updated his 8Gb FC versus 10Gb FCoE post, making it another must-read item for anyone interested in FC/FCoE debate. The fact that BRCDBreams decided to write a comment including “There is much dispute regarding the future market for FCoE vs. Fibre Channel” makes it even more fun to read.

Marko continues his myth debunking crusade. This week’s target: VLAN 1.

Canadian ISP going crazy – you probably know what my position on the Net Neutrality issue is, but this is plain stupid.

Yandy (the Packet Maniac) wrote a nice comparison of vPC and VSS features. You’ll find even more MLAG architectures in my blog and in my Data Center webinar (register or buy a recording).

I am positive Gartner told you a while ago that outsourcing will lower your overall costs. For whatever reason, some outsourcing projects turn out the same way as some multi-vendor networks.

2 comments:

  1. Re: the Canadian ISP - that's just nuts. Way overpriced. ISPs looking at implementing UBB should look at the plans available in Australia, where until recently, we have never had unlimited downloads.

    ReplyDelete
  2. Ivan,

    The multi vendor strategy isn't just about cost. I suspect that in some cases, having 2 players in a space can lower your cost because you can choose the ideal vendor for each installation and get better pricing. There are also other benefits, in that a critical flaw in 1 vendors code won't impact the other (see cisco long AS path handling flaws, various Junos FIB update delays in MX, etc.). If you have a big enough network that is going to be multivendor anyways to get best of breed different applications, you can split the fixed cost of working with that vendors OS over multiple products.

    For example, once you have vendor A in your edge switching, you already have staff with knowledge of their OS, have already integrated them into your various OSS systems and have a set of contacts there. If you add them to the vendor mix elsewhere, like core or PE, you can recycle a lot of that investment. It does however add some cost and time to your certification process, in that you now have to test new configs on two platforms.

    And perhaps the biggest advantage is that if one vendor makes a strategy change and drops your chosen product from their roadmap, you don't have 3-6 months of zero deployments while to work out the commerical and technical issues with a new supplier (or 3-6 months of buying dead-end hardware). You just switch your urgent deployments to vendor B while you replace vendor A's discontinued product.

    ReplyDelete

You don't have to log in to post a comment, but please do provide your real name/URL. Anonymous comments might get deleted.

Ivan Pepelnjak, CCIE#1354, is the chief technology advisor for NIL Data Communications. He has been designing and implementing large-scale data communications networks as well as teaching and writing books about advanced technologies since 1990. See his full profile, contact him or follow @ioshints on Twitter.