Archive

Category Archives for "Networking"

Sudan was cut off from the Internet for 25 days

Sudan was cut off from the Internet for 25 days
Sudan was cut off from the Internet for 25 days

Internet traffic started to come back in Sudan (with limitations) on Thursday, November 18, 2021. This happened after 25 days of an almost complete shutdown that affected the whole country. It’s a simple line going up on a chart for us, but for a country that also meant that Internet access was (at least in part) back on with all of what comes with it for businesses, communities, families and society as a whole.

You can see that trend on Cloudflare Radar, in particular after 13:00 UTC (15:00 local time). After that Internet traffic went up like we haven’t seen at all in the previous three weeks.

Sudan was cut off from the Internet for 25 days
Sudan was cut off from the Internet for 25 days

Internet access was mostly cut off on October 25, 2021, after a political turmoil in the country. A Sudanese court previously ordered the restoration of Internet access on November 9, but until last Thursday, November 18, there were no signs of services returning to normal. The biggest Internet access shutdown in recent history in the country was back in 2019 — for a full 36 days.

Looking back at the last 30 days Cloudflare Radar shows very distinctively a big difference from what was previously normal in the country.

Sudan was cut off from the Internet for 25 days

On Wednesday, Continue reading

Custom Groups and Deployment Templates in netsim-tools

Using custom templates to test IP anycast with MPLS was fun, but as I got into interesting discussions focusing on convoluted details, I found myself going through the same set of steps too many times.

It started with the need to specify individual devices in netlab config command to create new loopback interfaces on anycast servers but not on any other device in the lab. Wouldn’t it be nice to have a group of devices (similar to Ansible groups) that one could use in the limit parameter of netlab config?

netlab Custom Groups and Deployment Templates

Using custom templates to test IP anycast with MPLS was fun, but as I got into interesting discussions focusing on convoluted details, I found myself going through the same set of steps too many times.

It started with the need to specify individual devices in netlab config command to create new loopback interfaces on anycast servers but not on any other device in the lab. Wouldn’t it be nice to have a group of devices (similar to Ansible groups) that one could use in the limit parameter of netlab config?

IETF 112

Here the rest of the notes from some selected working group meetings that caught my attention at the recent IETF 112 meeting that are not related to DNS work.

Network Performance Update: Full Stack Week

Network Performance Update: Full Stack Week

This blog was published on November 20, 2021. As we continue to optimize our network we're publishing regular updates, which are available here.

Network Performance Update: Full Stack Week

A little over two months ago, we shared extensive benchmarking results of last mile networks all around the world. The results showed that on a range of tests (TCP connection time, time to first byte, time to last byte), and on a range of measurements (p95, mean), that Cloudflare was the fastest provider in 49% of networks around the world. Since then, we’ve worked to continuously improve performance until we’re the fastest everywhere. We set a goal to grow the number of networks where we’re the fastest by 10% every Innovation Week. We met that goal during Birthday Week (September 2021).

Today, we’re proud to report we blew the goal away for Full Stack Week (November 2021). Cloudflare measured our performance against the top 1,000 networks in the world (by number of IPv4 addresses advertised). Out of those, Cloudflare has become the fastest provider in 79 new networks, an increase of 14% of these 1,000 networks. Of course, we’re not done yet, but we wanted to share the latest results and explain how we did it.

However, Continue reading

FCC auction for prime 5G bandwidth rakes in $21.8B

The FCC auction for a prime band of 5G wireless spectrum has attracted $21.8 billion in bids, underscoring the importance of the frequency blocks that range from 3.5GHz to 3.55GHz.Of the 4,060 blocks available for discrete geographic areas throughout the country, all but 19 sold during the 29-day auction, according to the FCC.[Get regularly scheduled insights by signing up for Network World newsletters.] Carriers have been hungrily buying up this mid-range spectrum that some call the Goldilocks Zone because its transmissions propagate over a significant distance and also support high data rates. Earlier this year an auction for blocks of spectrum in the 3.7GHz (C-band) range raised more than $81 billion. To read this article in full, please click here

IP Class is Now in Session

You may have seen something making the rounds on Twitter this week about a couple of proposed drafts designed to alleviate the problems with IPv4 exhaustion by repurposing some old IP spaces that aren’t available for use right now. Specifically:

Ultimately, this is probably going to fail for a variety of reasons and looks like it’s more of a suggestion than anything else but I wanted to take a moment to talk about why this isn’t an effective way of fixing address issues.

Error Bearers

The first reason that the Schoen drafts are going to fail is because most of the operating systems in the world won’t allow you to use reserved spaces for a system address. Because we knew years ago that certain spaces were marked as non-usable the logic was configured into the system to disallow the use of those spaces. And even if the system isn’t configured to disallow that space there’s no guarantee the traffic is going to be transmitted.

Let’s take 127/8 as a good example. Was it a smart idea to mark 16 million addresses as loopback host-only space? Nope. But that ship has sailed and Continue reading

Xilinx launches a data-center accelerator for HPC

Xilinx has introduced its latest data-center accelerator, the Alveo U55C, which it says is its most powerful accelerator yet thanks to a memory change.For the most part, the FPGA-powered Alveo U55C is similar to its predecessor, Alveo U280. But the U280 has 8GB of HBM2 memory and 16GB of DDR4 DRAM, while the U55C comes with 16GB of HBM2 memory, and no DDR4. HBM2 is considerably faster and more expensive than DDR4 memory.[Get regularly scheduled insights by signing up for Network World newsletters.] By going to all HBM2 and removing the DDR4, Xilinx is able to increase performance and considerably reduce power and size. The Alveo U55C card is a single-slot full height, half length (FHHL) form factor vs. the full height, full length, dual width form of the U280. It also has a much lower power draw, 150W vs. 215W.To read this article in full, please click here

Xilinx launches a data-center accelerator for HPC

Xilinx has introduced its latest data-center accelerator, the Alveo U55C, which it says is its most powerful accelerator yet thanks to a memory change.For the most part, the FPGA-powered Alveo U55C is similar to its predecessor, Alveo U280. But the U280 has 8GB of HBM2 memory and 16GB of DDR4 DRAM, while the U55C comes with 16GB of HBM2 memory, and no DDR4. HBM2 is considerably faster and more expensive than DDR4 memory.[Get regularly scheduled insights by signing up for Network World newsletters.] By going to all HBM2 and removing the DDR4, Xilinx is able to increase performance and considerably reduce power and size. The Alveo U55C card is a single-slot full height, half length (FHHL) form factor vs. the full height, full length, dual width form of the U280. It also has a much lower power draw, 150W vs. 215W.To read this article in full, please click here