Road to gRPC

Road to gRPC
Road to gRPC

Cloudflare launched support for gRPC® during our 2020 Birthday Week. We’ve been humbled by the immense interest in the beta, and we’d like to thank everyone that has applied and tried out gRPC! In this post we’ll do a deep-dive into the technical details on how we implemented support.

What is gRPC?

gRPC is an open source RPC framework running over HTTP/2. RPC (remote procedure call) is a way for one machine to tell another machine to do something, rather than calling a local function in a library. RPC has been around in the history of distributed computing, with different implementations focusing on different areas, for a long time. What makes gRPC unique are the following characteristics:

  • It requires the modern HTTP/2 protocol for transport, which is now widely available.
  • A full client/server reference implementation, demo, and test suites are available as open source.
  • It does not specify a message format, although Protocol Buffers are the preferred serialization mechanism.
  • Both clients and servers can stream data, which avoids having to poll for new data or create new connections.

In terms of the protocol, gRPC uses HTTP/2 frames extensively: requests and responses look very similar to a normal HTTP/2 request.

Continue reading

Network Break 307: Cisco Launches Catalyst 8000 Edge Routers; Juniper Spends $450 Million For 128 Technology

Today's Network Break covers new SD-WAN routers from Cisco, a giant Juniper acquisition, new WIPS capabilities from Extreme, the United States' anti-trust lawsuit against Google, Intel selling its NAND business, and Space Networking!

The post Network Break 307: Cisco Launches Catalyst 8000 Edge Routers; Juniper Spends $450 Million For 128 Technology appeared first on Packet Pushers.

Low demand and high production create a bargain for SSD drives

Analysts had expected 2020 to be a year of rising NAND flash prices after a supply glut in 2019. The reasoning, backed by years of repeating patterns, was that after an oversupply, vendors like SK Hynix and Micron would slow production to drive prices up.Cue COVID-19 and the ensuing chaos.TrendForce, a market research firm that follows the memory market, said there has been a general decline in contract prices starting last quarter due to oversupply. This oversupply situation is attributed to the accumulation of inventory caused by the economic fallout from the COVID-19 pandemic. Learn more:To read this article in full, please click here

Low demand and high production create a bargain for SSD drives

Analysts had expected 2020 to be a year of rising NAND flash prices after a supply glut in 2019. The reasoning, backed by years of repeating patterns, was that after an oversupply, vendors like SK Hynix and Micron would slow production to drive prices up.Cue COVID-19 and the ensuing chaos.TrendForce, a market research firm that follows the memory market, said there has been a general decline in contract prices starting last quarter due to oversupply. This oversupply situation is attributed to the accumulation of inventory caused by the economic fallout from the COVID-19 pandemic. Learn more:To read this article in full, please click here

The Week in Internet News: U.S. Files Antitrust Case Against Google

"In the news" text on yellow background

Searching for a monopoly: The U.S. Department of Justice has filed an antitrust lawsuit against Google, accusing the tech giant of illegal monopolies in search and search advertising, CNet reports. The DOJ has accused Google of acting as an Internet “gatekeeper.” Google disputed the allegations, saying people use its services because they choose to, not because they have to.

New networking: The Lower Brule Sioux Tribe in central South Dakota has advanced a plan to provide computers and high-speed Internet connections to all students and teachers during the COVID-19 pandemic, the Argus Leader says. Since early June, the tribe has been working on a plan to build its own wireless Internet network, intended to cover the 207-square-mile Lower Brule reservation. The new network is the first-of-its-kind in South Dakota, and it began limited operations in July.

A bumpy relaunch: The French government’s relaunch of its COVID-19 tracing app, called, “TousAntiCovid,” hit some snags when it was downloaded more than 500,000 times in the hours following its launch, the BBC says. The traffic led to some stability problems, with some people unable to launch the app.

DNS vs. crime: Securing the Internet’s domain name system is a crucial step in Continue reading

New on ipSpace.net: Graph Algorithms

After a bit more than a year, we ran another math-focused webinar last week: Rachel Traylor came back to talk about graph algorithms, focusing on tree-, path- and center problems.

In her lecture you’ll find:

  • maximum branching algorithms (and I couldn’t stop wondering why we don’t use them for OSPF- or IS-IS flooding)
  • path algorithms including the ones used in OSPF, IS-IS, or BGP, as well as algorithms that find K shortest paths
  • center problems (for example: where do I put my streaming server or my BGP route reflector)

You’ll need Standard or Expert ipSpace.net subscription to watch the videos.

New on ipSpace.net: Graph Algorithms

After a bit more than a year we ran another math-focused webinar last week: Rachel Traylor came back to talk about graph algorithms, focusing on tree-, path- and center problems.

In her lecture you’ll find:

  • maximum branching algorithms (and I couldn’t stop wondering why we don’t use them for OSPF- or IS-IS flooding)
  • path algorithms including the ones used in OSPF, IS-IS, or BGP, as well as algorithms that find K shortest paths
  • center problems (for example: where do I put my streaming server or my BGP route reflector)

You’ll need Standard or Expert ipSpace.net subscription to watch the videos.

No Rush

Intro

We often treat our careers like it’s a race. With only a winner. We setup goals where we want to get a degree by a certain age. Get that certification at another age. Get that job at a certain age and we judge our success by if we make more than say 100k per year. Because that’s what we’ve been told.

However, building a successful career in IT is nothing like that.

Stress

I’ve been there myself and felt the stress. I started my university studies when I was 22. I felt old at the time when I was surrounded by people that were 18-19 years old. I know that people where I lived before my university studies had started asking questions of the kind if I wasn’t to become anything. To do something with my life. I needed a few years break from school before going to university studies , and it turns out that was a great decision. I was able to study in a matter I had never done before.

One of the goals I setup in my career was to become a CCIE by 30. I’m not sure why. It just seemed like getting it Continue reading

Helios: hyperscale indexing for the cloud & edge – part 1

Helios: hyperscale indexing for the cloud & edge, Potharaju et al., PVLDB’20

On the surface this is a paper about fast data ingestion from high-volume streams, with indexing to support efficient querying. As a production system within Microsoft capturing around a quadrillion events and indexing 16 trillion search keys per day it would be interesting in its own right, but there’s a lot more to it than that. Helios also serves as a reference architecture for how Microsoft envisions its next generation of distributed big-data processing systems being built. These two narratives of reference architecture and ingestion/indexing system are interwoven throughout the paper. I’m going to tackle the paper in two parts, focusing today on the reference architecture, and in the next post on the details of Helios itself. What follows is a discussion of where big data systems might be heading, heavily inspired by the remarks in this paper, but with several of my own thoughts mixed in. If there’s something you disagree with, blame me first!

Why do we need a new reference architecture?

Cloud-native systems represent by far the largest, most distributed, computing systems in our history. And the established cloud-native architectural principles behind them Continue reading

Why Biden: Principle over Party

There exist many #NeverTrump Republicans who agree that while Trump would best achieve their Party's policies, that he must nonetheless be opposed on Principle. The Principle at question isn't about character flaws, such as being a liar, a misogynist, or a racist. The Principle isn't about political policies, such as how to hand the coronavirus pandemic, or the policies Democrats want. Instead, the Principle is that he's a populist autocrat who is eroding our liberal institutions ("liberal" as in the classic sense).

Countries don't fail when there's a leftward shift in government policies. Many prosperous, peaceful European countries are to the left of Biden. What makes prosperous countries fail is when civic institutions break down, when a party or dear leader starts ruling by decree, such as in the European countries of Russia or Hungary.

Our system of government is like football. While the teams (parties) compete vigorously against each other, they largely respect the rules of the game, both written and unwritten traditions. They respect each other -- while doing their best to win (according to the rules), they nonetheless shake hands at the end of the match, and agree that their opponents are legitimate.

The rules of the Continue reading

ACI Fabric Access Policies Part 4: Leaf Interface Profile, Leaf Switch Policy Group, and Leaf Switch Profile,


Leaf Interface Profile

 

This section explains how to create an object Interface Profile whose basic purpose is to attach the set of physical interfaces into this object. Phase 6 in Figure 1-40 illustrates the APIC Management Information Model (MIM) from the Interface Profile perspective. We are adding an object L101__102_IPR under the class AccPortP (Leaf Interface Profile). The name of the object includes Leaf switch identifiers (Leaf-101 and Leaf-102) in which I am going to use this Interface Profile. This object has a Child object Eth1_1-5 (class InfraHPorts) that defines the internet block and which has a relationship with the object Port_Std_ESXi-Host_IPG. By doing this we state that ethernet interfaces 1/1-5 are LLDP enabled 10Gbps ports which can use VLAN Identifiers from 300-399. Note that in this phase we haven’t yet specified in which switches we are using this Interface Profile.

 The RN rules used with related objects:

 Objects created under the class InfraAccportP (Leaf Interface Profile):Prefix1-{name}, where the Prefix1 is “accportprof”. This gives us RN “accportprof-L101_L102_IPR”.

 Objects created under the class InfraHPortS (Access Port Selector): Prefix1-{name}-Prefix2-{type}, where the Prefix1 is “hports” and the Prefix2 is “typ”. This gives us RN “hports-Eth1_1-5_typ-range”.

Objects created under the class InfraPortBlk (Access Port Block): Prefix1-{name}, where the Prefix1 is “portblk” and where the name is Property (autogenerated). This gives us the RN “portblk-Block2”.



Figure 1-39: APIC MIM Reference: Interface Profile.

Continue reading

ACI Fabric Access Policies Part 3: AAEP, Interface Policy and Interface Policy Group

 

Attachable Access Entity Profile - AAEP


This section explains how to create an object Attachable Access Entity Profile (AAEP) that is used for attaching a Domain into Port Group. Phase 3 in Figure 1-20 illustrates the APIC Management Information Model (MIM) from the AAEP perspective. Class AttEntityP is a Child class for infra, and they both belong to packages Infra. I have already added the object attentp-AEP_PHY into the figure.The format of the RN for this object is Prefix1-{name}, where the Prefix1 is attentp. This gives us the RN attentp-PHY-AEP.



Figure 1-20: APIC MIM Reference: Attachment Access Entity Profile.

Continue reading

Worth Reading: The Shared Irresponsibility Model in the Cloud

A long while ago I wrote a blog post along the lines of “it’s ridiculous to allow developers to deploy directly to a public cloud while burdening them with all sorts of crazy barriers when deploying to an on-premises infrastructure,” effectively arguing for self-service approach to on-premises deployments.

Not surprisingly, the reality is grimmer than I expected (I’m appalled at how optimistic my predictions are even though I always come across as a die-hard grumpy pessimist), as explained in The Shared Irresponsibility Model in the Cloud by Dan Hubbard.

For more technical details, watch cloud-focused ipSpace.net webinars, in particular the Cloud Security one.

1 … 863 864 865 866 867 … 3,840