Latency: the Killer of Spread-Out Application Stack Ideas

A few months ago I described how bandwidth limitations shatter the dreams of spread-out application stacks with elements residing (or being dynamically migrated) between data centers. Today let’s focus on bandwidth’s ugly cousin: latency.

TL&DR Summary: Spreading the server components of an application across multiple locations (multiple data centers or hybrid cloud deployments) can easily result in dismal performance even when there’s plenty of bandwidth available.

How Does MPLS-TE Interact with QoS

MPLS Traffic Engineer is sometimes promoted as a QoS solution (it seems bandwidth calendaring is a permanent obsession of some networking engineers, and OpenFlow is no more a solution than MPLS-TE was ;), but in reality it’s pretty hard to make the two work together seamlessly (just ask anyone who had to implement auto-bandwidth MPLS-TE in a large network).

Not surprisingly, we addressed the topic during our MPLS Tech Talk.

Totally Redesigned Free Content Web Site

One of the main complaints I was continuously getting about my free content is that there’s simply too much of it, and that it’s impossible to find what one is looking for.

New Year holidays gave me enough time to implement a project that has been on my to-do list for almost a year: total redesign of the free content web site. Feedback highly appreciated!

BGP Deaggregation with Conditional Route Injection

Whenever there’s a weird request to do something totally illogical with BGP, there’s a knob in Cisco IOS to get it done (and increase the heartburn of CCIE candidates). Conditional Route Injection (the ability to insert more specific prefixes into BGP without having them in the IP routing table) is one of them.

Keep in mind: being a MacGyver is not a long-term strategy. Just because you can doesn’t mean that you should.

That’s It for 2014

A dozen webinars, tens of public presentations and on-site workshops, numerous highly interesting ExpertExpress sessions, three books and over 250 blog posts. That should be enough for a year; it’s time to go offline.

I hope your company has a New Year freeze (and not let’s upgrade everything over New Year policy), so you’ll be able to do the same and enjoy some time during the rest of the year with your loved ones. See you in 2015!

MPLS Tech Talks: RSVP-TE 101

After discussing the basics of MPLS, MPLS-TE and LDP, and the relationship between FECs, LDP and BGP, Seamus and myself focused on another interesting topic: how MPLS protocol stack uses RSVP to implement traffic engineering.

VRF Lite on Nexus 5600

One of the networking engineers using my ExpertExpress to validate their network design had an interesting problem: he was building a multi-tenant VLAN-based private cloud architecture with each tenant having multiple subnets, and wanted to route within the tenant network as close to the VMs as possible (in the ToR switch).

He was using Nexus 5600 as the ToR switch, and although there’s conflicting information on the number of VRFs supported by that switch (verified topology: 25 VRFs, verified maximum: 1000 VRFs, configuration guide: 64 VRFs), he thought 25 VRFs (tenant routing domains) might be enough.

Just Published: Scaling Overlay Virtual Networking Videos

The edited videos for Scaling Overlay Virtual Networking webinar are available on ipSpace.net Content site. Nuage Networks sponsored the webinar; the videos are thus publicly available (without registration).

Webinars in 2014, and a Quick Peek Into 2015

I promise engineers who renew their subscription 4-6 new webinars a year. It’s time to see whether I kept that promise in 2014.

TL&DR summary: it was a great year, but I still missed a few things.

L2VPN over IPv6 with Snabb Switch on Software Gone Wild

Highly customizable high-speed virtual switch written in Lua sounds great, but is it really that easy to use? Simon Leinen was kind enough to get me in touch with Alex Gall, his colleague at Switch, who's working on an interesting project: implementing L2VPN over IPv6 with Snabb Switch.

Facebook Next-Generation Fabric

Facebook published their next-generation data center architecture a few weeks ago, resulting in the expected “revolutionary approach to data center fabrics” echoes from the industry press and blogosphere.

In reality, they did a great engineering job using an interesting twist on pretty traditional multi-stage leaf-and-spine (or folded Clos) architecture.

Performance Tests and Out-of-Box Performance

Simonp made a perfectly valid point in a comment to my latest OVS blog post:

Obviously the page you're referring to is a quick-and-dirty benchmark. If you wanted the optimal numbers, you would have to tune quite a few parameters just like for hardware benchmarks (sysctl kernel parameters, Jumbo frames, ...).

While he’s absolutely right, this is not the performance data a typical user should be looking for.

Last Call: Overlay Virtual Networks in Software Defined Data Centers

If you want to get a free copy of my Overlay Virtual Networks in Software-Defined Data Centers book, download it now. The offer will expire by December 15th.

Just published: Enterprise IPv6 videos

The edited videos for my Enterprise IPv6 webinars have been published on my.ipspace.net. Enjoy!

Load Balancing in Google Network

Todd Hoff (of the HighScalability fame) sent me a link to an interesting video describing load-balancing mechanisms used at Google and how they evolved over time.

If the rest of the blog post feels like Latin, you SHOULD watch the Load Balancing and Scale-Out Application Architecture webinar.

The beginning of the story resembles traditional enterprise solutions:

Scaling Distributed Systems Is Hard

Stumbled upon a hilarious description of challenges encountered when trying to scale distributed systems (cluster of controllers running centralized control plane comes to mind).

It starts with “If someone tells you that scaling out a distributed system is easy they are either lying or drunk, and possibly both,” and gets better and better. Enjoy!

Hotel California Effects of Public Clouds

In his The Case for Hybrids blog post Mat Mathews described the Hotel California effect of public clouds as: “One of the most oft mentioned issues with public cloud is the difficulty in getting out.” Once you start relying on cloud provider APIs to provide DNS, load balancing, CDN, content hosting, security groups, and a plethora of other services, it’s impossible to get out.

Interestingly, the side effects of public cloud deployments extend into the realm of application programming, as I was surprised to find out during one of my Expert Express engagements.

Should I Really Program My Network?

In my presentation @ SDN Meetup in Stockholm, I tried to answer a simple question: “Should I really program my network?” and obviously had to start with an even simpler one: “What is SDN?”

The video of the presentation is already available on YouTube, and you can watch the slides on my content web site.

Also, make sure you watch other presentations from that event, particularly David Barroso’s SDN Internet Router.

MPLS P-Router, Router or Layer-3 Switch?

One of my readers is struggling with the aftermath of marketing gimmicks:

We will be implementing a new network soon, and we're discussing P-routers versus regular routers versus switches. I'm looking for arguments to go one way or the other.

TL&DR: there’s no difference between router and L3 switch.

Can You AS-Prepend a Single Host Route?

Someone recently sent me this question:

Is it possible to prepend one IP address from a public IPv4 segment?

I don’t want to know what crazy stunt this engineer was forced to pull off, but just in case you land in a similar quandary here’s how you shoelace yourself out of it.