Measuring network quality to better understand the end-user experience

You’re visiting your family for the holidays and you connect to the WiFi, and then notice Netflix isn’t loading as fast as it normally does. You go to speed.cloudflare.com, fast.com, speedtest.net, or type “speed test” into Google Chrome to figure out if there is a problem with your Internet connection, and get something that looks like this:

If you want to see what that looks like for you, try it yourself here. But what do those numbers mean? How do those numbers relate to whether or not your Netflix isn’t loading or any of the other common use cases: playing games or audio/video chat with your friends and loved ones? Even network engineers find that speed tests are difficult to relate to the user experience of… using the Internet..

Amazingly, speed tests have barely changed in nearly two decades, even though the way we use the Internet has changed a lot. With so many more people on the Internet, the gaps between speed tests and the user’s experience of network quality are growing. The problem is so important that the Internet’s standards organization is paying attention, too.

From a high-level, there are three grand network Continue reading

IPv6 Addressing on Point-to-Point Links

One of my readers sent me this question:

In your observations on IPv6 assignments, what are common point-to-point IPv6 interfaces on routers? I know it always depends, but I’m hearing /64, /112, /126 and these opinions are causing some passionate debate.

(Checks the calendar) It’s 2023, IPv6 RFC has been published almost 25 years ago, and there are still people debating this stuff and confusing those who want to deploy IPv6? No wonder we’re not getting it deployed in enterprise networks ;)

IPv6 Addressing on Point-to-Point Links

One of my readers sent me this question:

In your observations on IPv6 assignments, what are common point-to-point IPv6 interfaces on routers? I know it always depends, but I’m hearing /64, /112, /126 and these opinions are causing some passionate debate.

(Checks the calendar) It’s 2023, IPv6 RFC has been published almost 25 years ago, and there are still people debating this stuff and confusing those who want to deploy IPv6? No wonder we’re not getting it deployed in enterprise networks ;)

Tracing function calls

Sometimes you want to see functions of a library, as they’re called. I know of two ways of doing this.

Let’s have a super simple test program:

#include<iostream>
#include<unistd.h>

void func1() {}
void func2() {}

int main()
{
  std::cout << "Hello world\n";
  func1();
  func2();
  func1();

  // Wait a bit for bpftrace to be able to aquire the function name.
  // Not applicable for something that doesn't exist.
  sleep(1);
}

bpftrace

Start a bpftrace in one terminal, and run the program in another.

$ sudo bpftrace -e 'uprobe:./a.out:func* { print(func); }'
Attaching 2 probes...
func1()
func2()
func1()

GDB

$ gdb a.out
[…]
(gdb) rbreak func.*
[…]
(gdb) commands
Type commands for breakpoint(s) 1-3, one per line.
End with a line saying just "end".
>silent
>bt 1
>cont
>end
(gdb) r
Starting program: […]/a.out
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".
#0  0x0000555555555215 in _GLOBAL__sub_I__Z5func1v ()
Hello world
#0  0x000055555555516d in func1() ()
#0  0x0000555555555174 in func2() ()
#0  0x000055555555516d in func1() ()
[Inferior 1 (process 424744) exited normally]
(gdb)

Which to use?

bpftrace is lower (but Continue reading

What’s the Enterprise Impact of Microsoft’s Open-Sourced SONiC?

How to create netstat aliases to help focus on network activity

The netstat command provides a tremendous amount on information on network activity. With the -s option (netstat -s), it will display summaries for various protocols such as packets received, active connections, failed connections and a lot more. While the data is extensive enough to make you dizzy, the more you get used to what the command's output looks like, the more you'll become familiar with what to expect and maybe even get better at spotting what's unusual. In this post, we're going to look at various portions of the netstat -s command's output using crafted aliases to make it easier.To read this article in full, please click here

How to create netstat aliases to help focus on network activity

The netstat command provides a tremendous amount on information on network activity. With the -s option (netstat -s), it will display summaries for various protocols such as packets received, active connections, failed connections and a lot more. While the data is extensive enough to make you dizzy, the more you get used to what the command's output looks like, the more you'll become familiar with what to expect and maybe even get better at spotting what's unusual. In this post, we're going to look at various portions of the netstat -s command's output using crafted aliases to make it easier.To read this article in full, please click here

Network Break 426: NetBox Labs Raises $20 Million; Intel Foundry, Arm Team Up To Make SoCs

Network Break 426: NetBox Labs Raises $20 Million; Intel Foundry, Arm Team Up To Make SoCs

This week's Network Break discusses NetBox Labs's $20 million funding round and why it spun itself out of NS1's IBM acquisition, Intel Foundry and Arm teaming up on SoC manufacturing, why Amazon sees headwinds for public cloud spending (and why the opportunity is still huge), how the juice-jacking threat got hyped, and more tech news.

The post Network Break 426: NetBox Labs Raises $20 Million; Intel Foundry, Arm Team Up To Make SoCs appeared first on Packet Pushers.

How to Take a Wireless-First Approach to Enterprise Network Design

The Future of SD-WAN and Fixed Cellular Connections

Intel news roundup: chiplets milestone, server exit, and ARM deal

Intel had a busy week. A trio of news announcements revealed its chiplets progress, a manufacturing agreement with Arm, and the shedding of another non-core line of business.Prototype multi-die chips heading to DoD The biggest news is that Intel has begun to ship prototype multi-die chips to the U.S. Department of Defense more than a year ahead of schedule. The DoD project known as State-of-the-Art Heterogeneous Integrated Packaging (SHIP) is an ambitious plan that will connect Intel’s CPUs, FPGAs, ASICs and government-developed chiplets all within the same processor packaging, as opposed to multiple separate dies.AMD was the first to pursue the chiplet design, but AMD took a different approach in that it broke up large, monolithic CPUs into smaller chips. So, instead of one physical piece of silicon with 32 cores, it created four chiplets with eight cores each connected by high-speed interconnects. The idea is that it’s much easier to manufacture an eight-core chip than a 32-core chip.To read this article in full, please click here

Intel news roundup: chiplets milestone, server exit, and ARM deal

Intel had a busy week. A trio of news announcements revealed its chiplets progress, a manufacturing agreement with Arm, and the shedding of another non-core line of business.Prototype multi-die chips heading to DoD The biggest news is that Intel has begun to ship prototype multi-die chips to the U.S. Department of Defense more than a year ahead of schedule. The DoD project known as State-of-the-Art Heterogeneous Integrated Packaging (SHIP) is an ambitious plan that will connect Intel’s CPUs, FPGAs, ASICs and government-developed chiplets all within the same processor packaging, as opposed to multiple separate dies.AMD was the first to pursue the chiplet design, but AMD took a different approach in that it broke up large, monolithic CPUs into smaller chips. So, instead of one physical piece of silicon with 32 cores, it created four chiplets with eight cores each connected by high-speed interconnects. The idea is that it’s much easier to manufacture an eight-core chip than a 32-core chip.To read this article in full, please click here

AWS and Alexa services back online after outage

Amazon web services and Amazon voice assistant Alexa were back online after an outage on Sunday that lasted over three hours. Users also reported issues with accessing the Amazon mobile app.Downdetector, the website that tracks outages, reported that Alexa was down for thousands of users in the US, while an outage in AWS was reported by hundreds of users. More than 16,000 users reported an outage in Alexa at the peak of the disruption.Users could not complete the account signup process and received error messages on their billing console.The billing console is used to manage ongoing payments and payment methods registered to AWS accounts.To read this article in full, please click here

AWS and Alexa services back online after outage

Amazon web services and Amazon voice assistant Alexa were back online after an outage on Sunday that lasted over three hours. Users also reported issues with accessing the Amazon mobile app.Downdetector, the website that tracks outages reported that Alexa was down for thousands of users in the United States, while an outage in AWS was reported by hundreds of users. More than 16,000 users reported an outage in Alexa at the peak of the disruption.Users could not complete the account signup process and received error messages on their billing console.The billing console is used to manage ongoing payments and payment methods registered to AWS accounts.To read this article in full, please click here

Using VLAN and VRF Links in netlab Topologies

I already mentioned the introduction of VRF- and VLAN access links in netlab release 1.5.1. Let’s see how they can simplify your lab topologies.

I always tried to make lab topologies as concise as I could, sometimes cheating using JSON-in-YAML syntax. For example, the topology describing three routers running OSPF could be as simple as this:

module: [ ospf ]
nodes: [ r1, r2, r3 ]
links: [ r1-r2, r2-r3, r3-r1 ]

Let’s unravel that:

Using VLAN and VRF Links in netlab Topologies

I already mentioned the introduction of VRF- and VLAN access links in netlab release 1.5.1. Let’s see how they can simplify your lab topologies.

I always tried to make lab topologies as concise as I could, sometimes cheating using JSON-in-YAML syntax. For example, the topology describing three routers running OSPF could be as simple as this:

module: [ ospf ]
nodes: [ r1, r2, r3 ]
links: [ r1-r2, r2-r3, r3-r1 ]

Let’s unravel that:

Enforcing First AS in BGP

The BGP RFCs state that external BGP peers should insert their own AS into the AS PATH advertised to eBGP peers. Some peers strip their AS, generally for commercial gain. Juniper and Cisco have opposite default behaviors for handling this. Make sure you set bgp enforce-first-as on Juniper routers. Caveats apply.

Background: Traffic Anomalies

A few years ago I was looking at some traffic reporting anomalies. My IPFIX data said that traffic with next-hop AS <dodgy-AS> was around 3Gb. But my SNMP data showed that a PNI to that peer was doing 8-10Gb.

I first doubted my router, because I had issues with IPFIX in the past on that specific platform. I also wondered about sampling rates. I have high flow rates, and need to set the sampling to be more coarse. But it was a big anomaly.

Slicing & dicing the data different ways, and chatting to colleagues about it, we saw what was going on. IPFIX showed the right volumes when reporting on destination interface. But some prefixes received from the peer did not contain the peer’s AS. We still accepted them.

Huh? Isn’t it normal behavior, to insert your own AS into any prefixes you advertise to Continue reading

ChatGPT your Project Docs

I mentioned in my last post that I would start updating practical use cases for generative AI. The first example use case is your project documentation. As software engineers, we spend days of our lives writing project docs. The reality is that people learn and consume data differently. Overly detailed docs may cause some people’s eyes to glaze over while ...

The post ChatGPT your Project Docs appeared first on NetworkStatic | Brent Salisbury's Blog.

RISC-V optimization and -mtune

I’ve been getting into RISC-V optimization recently. Partly because I got my SiFive VisionFive 2, and partly because unlike x86 the number of RISC-V instructions is so managable that I may actually have a chance at beating the compiler.

I’m optimizing the inner loops of GNURadio, or in other words the volk library. I’ve been getting up to a about a doubling of the speed compared to the compiled C code, depending on the function.

But it got me thinking how far I could tweak the compiler and its options, too.

Yes, I should have done this much sooner.

Many years ago now I built some data processing thing in C++, and thought it ran too slowly. Sure, I did a debug build, but how much slower could that be? Half speed? Nope. 20x slower.

Of course this time I never compared to a debug build, so don’t expect that kind of difference. Don’t expect that it’ll reach my hand optimized assembly either, imperfect as it may be.

The test code

This may look like a synthetic benchmark, in simplified C++:

complex volk_32fc_x2_dot_prod_32fc_generic(const vector<complex> &in1,
                                           const vector<complex> &in2)
{
  complex res;
  for (unsigned int i = 0; i  Continue reading