Vlad Krasnov recently joined CloudFlare to work on low level optimization of CloudFlare's servers. This is the first of a number of blog posts that will include code he's optimized and open sourced.
In a recent post, Kazuho's Weblog describes an improvement to PicoHTTPParser. This improvement utilizes the SSE4.2 instruction PCMPESTRI in order to find the delimiters in a HTTP request/response and parse them accordingly. This update, compared to the previous version of the code, is impressive.
CC BY-SA 2.0 image by Intel Free Press
PCMPESTRI is a versatile instruction that allows scanning of up to 16 bytes at once for occurrences of up to 16 distinct characters (bytes), or up to 8 ranges of characters (bytes). It can also be used for string and substring comparison. However, there are a few drawbacks: the instruction has a high latency of 11 cycles, and is limited to 16 bytes per instruction. It's also under utilized for range comparison in PicoHTTPParser, because it only tests two or three ranges per invocation (out of eight it is capable of). Furthermore, some simple math (16 bytes / 11 cycles) shows that using this instruction limits the parser to 1.45 bytes/cycle throughput.
As mentioned in an earlier post GNS3 is moving ahead fast. Currently at version 1.2.1 the GNS3 is looking great. Compared with the version 1.0 Beta 1 which I had installed, the 1.2.1 is not only more stable, but it has the Menu more clean and compact. For example now there is only one Preferences menu where you can adjust all your settings.
Big thinkers think about giving, doing, and ideas. Small thinkers think about getting and having. If you’re comparing to or gossiping about others — you’re not thinking at all.
Nearly two years ago, we broke the story about the activation of the first submarine cable connecting Cuba to the global Internet – a cable that, prior to its activation in January 2013, mysteriously lay dormant on the ocean floor for nearly two years. When the Cuban government issued a confirmation in the days following our report, it contained the following statement:
|
![]() |
In other words, Cubans should not expect greater access to the Internet just because the ALBA-1 submarine cable was now in operation. Yesterday’s historic agreement to begin normalizing relations between Cuba and the United States contains a pledge by the Cuban government to “greatly expand its citizens’ access to the Internet.” What exactly this pledge entails will determine how the Internet evolves in Cuba in the near term. Decision makers in Cuba should look at another country that recently opened up its telecom sector and is presently experiencing an explosion in Internet growth: Myanmar.
Cuban Isolation
The isolation of Cuba is plainly evident when looking at a map of the submarine cables in the Continue reading
A while ago I received the following question:
“Why I’m not seeing the prefixes received from the BGP peer when using the show ip bgp neighbors x.x.x.x received-routes while the soft-reconfiguration inbound is not enabled?”
Read more on Cisco BGP soft-reconfiguration and received-routes relation…
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.
The Alcatel-Lucent virtualized Simulator (vSim) is a virtualization-ready version of SR OS called SR OS-VM. This new operating system is designed to run in a virtual machine (VM) on a generic Intel x86 server. In control and management plane aspects, the vSim is functionally and operationally equivalent to an Alcatel-Lucent hardware-based SR OS router.The vSim is intended to be used as a laboratory tool to fully simulate the control and management plane of an SR OS node. The vSim is not intended to be used in a production network environment and the forwarding plane is limited to 250 pps per interface. Furthermore, without a license file it will run for 1 hour before reloading.
Host Software and Hardware Requirements
Virtual Machines Software and Hardware Requirements
Extract image from the zip file.
$ unzip TiMOS-SR-12.0.R6-vm.zip
$ cd vm/7xxx-i386/
Now a virtual disk sros-vm.qcow2 is extracted. To start Qemu virtual Continue reading
(yes, I know, it’s been a while… But it’s time to get back to this series) Up to this point in this series, we’ve been discussing the more technical aspects of how the Internet really works. Now I want to shift gears a little, and talk about some of the more political aspects — standards […]
The post HTIRW: Standards Bodies appeared first on Packet Pushers Podcast and was written by Russ White.
You might have noticed that blog content has been a bit sparse over the last few weeks. The reason I haven’t generated any new content is because all my spare time is taken up with preparing to migrate this site to a new hosting platform.
Sometime over the holiday season, I’ll be migrating this site from a hosted WordPress installation to Jekyll running on GitHub Pages. Given that I have 9 years of content (over 1,600 blog posts), this is a pretty fair amount of work.
Most of the “structural” work on the new site is already complete; you can get a preview of the site by visiting http://lowescott.github.io. There’s no content there yet (other than some boilerplate content), but you’ll be able to get a feel for how the new layout will look and work. As you can see, I’ll be using the Lanyon theme, which provides a nice clean layout and a good mobile as well as desktop experience.
There’s still some additional “structural” work to be done, such as adding support for comments (which will be handled via Disqus), but I hope to have that done in the next few days.
Once the Continue reading
We were proud to have DualSpark join us for a great webinar this week on automation on AWS using Ansible. Presenting from Ansible was Dave Johnson and Patrick McClory handled the discussion from the DualSpark side.
Ansible Automation on AWS: Best Practices by Battle-Hardened Experts
This week are talking about the IETF and it's inability to cope with massive change in networking around SDN and NFV. For example, there are more than 70 drafts on NETCONF models for common networking tasks that often overlap or repeat the same work. What does this means for standards development ?
The post Show 217 – IETF, YANG Proliferation and the Lack of Cooperation and Co-ordination appeared first on Packet Pushers Podcast and was written by Greg Ferro.
It is common to allocate /24 or /22 subnets to a single VLAN but William writes to ask why and whether is related to broadcasts. What is the best subnet size for VLAN allocation and why ? The answer isn't what you think.
The post How Many Hosts In An VLAN or IP Subnet and Why ? appeared first on EtherealMind.
Recently Arista released a white paper surrounding the idea that having deeper buffers running within the network can help to alleviate the incast congestion patterns that can present when a large number of many-to-one connections are happening within a network. Also known as the TCP incast problem. They pointedly targeted Hadoop clusters, as the incast problem can rear its ugly head when utilizing the Hadoop Cluster for MapReduce functions. The study used an example of 20 servers hanging off of a single ToR switch that has 40Gbps of uplink capacity within a Leaf/Spine network, presenting a 5:1 oversubscription ratio. This type of oversubscription was just seen in the recent release of the Facebook network that is used within their data centers. So its safe to assume that these types of oversubscription ratios are seen in the wild. I know I’ve run my fair share of oversubscribed networks in the past.
This particular study actually prods at what is the achilles heel of the traditional leaf/spine network design. All nodes being within 3 switch hops, (ToR <-> Spine <-> ToR), does provide a predictable pathing within the minds of the network operators of today, but I posit that Continue reading