Cloud Native Computing Foundation KubeCon + CloudNativeCon Europe 2017 (Berlin), 6 Apr 2017

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: Keynote: OpenTracing Project Update - Priyanka Sharma, LightStep

Description

Keynote: OpenTracing Project Update - Priyanka Sharma, LightStep

The OpenTracing project turned one year old a month ago. In the last year, it has been solidified as the API standard for getting production data out of complex, distributed systems. Come learn about the latest API updates, language and other OSS bindings, and see a demo that involves donuts.

About Priyanka Sharma
Priyanka Sharma is an entrepreneur with a passion for building developer products and growing them through open source communities. She heads Marketing and Partnerships at LightStep and also works on the OpenTracing project, an instrumentation standard for distributed tracing. In her copious spare time she advises startups at HeavyBit industries, an accelerator for developer products. Priyanka cofounded WakaTime, an open source time-tracker for developers and holds a BA in Political Science from Stanford University.
Join us for KubeCon + CloudNativeCon in Barcelona May 20 - 23, Shanghai June 24 - 26, and San Diego November 18 - 21! Learn more at https://kubecon.io. The conference features presentations from developers and end users of Kubernetes, Prometheus, Envoy and all of the other CNCF-hosted projects.

A

Please give a warm welcome to priyanka sharma, representing the open tracing projects.

A

Hi everyone, as Michelle, said I'm, Priyanka and I work on the open tracing project. A lot has happened for open tracing in the last year. It's become a member of CN CF for one thing, but I only have five minutes today, so I'm going to cut straight to the chase. Well, so open tracing just turn one a little before this conference last year, Ben siegelman a flight step, Adrian Cole of Zipkin and other folks, who spent a lot of time.

A

Thinking about distributed tracing came together and built a specification that can help companies individual developers get started with tracing much easier.

A

So companies, big and small, have started relying on open tracing two-port data from their system into a tracing solution of their choice. It could be a vendor or it could be something homegrown because of the vendor neutrality and portability of open tracing. This has been easy for them, also open source projects. Some of your favorites, probably as well as vendors in the tracing landscape, have joined in so some of the big ones recently that I've seen are spring for Java G RPC. We also have a ruby support now and on the vendor side.

A

Many are offering integrations for open tracing data models, so think of Zipkin instant, our Red Hat ocular from Red Hat at Sarah. So you might wonder why.

A

Why are all these people coming together and getting on the open tracing bandwagon? It's only a year old. After all, it's very simple distributed systems are killing engineers.

A

Many have died, okay, I'm just kidding, they are killing engineers productivity, though in the past, in the monolithic architecture, you have full control and visibility of this one large monolith, and you can see exactly what's going on. If a problem happens, you can pinpoint directly to it solve it. Move on, however, as systems get more distributed, more fragmented, those that visibility is gone and suddenly we're seeing mean time to resolution skyrocket and things like root, cause analysis really really painful, let alone root cause analysis at 3:00 a.m. in the morning for the pager carrying warrior.

A

So now you probably understand why I have this very intense graphic over here. It's almost killing people but distributed tracing saves the day, so this technology has been out there for a while now as exampled by dapper of google, where a company has been able to utilize traces to understand what's happening in a very complex system.

A

However, companies have not been able to take advantage of it because it's a high effort endeavor and in the past there was no standard that helped people know that all the effort they put into seeing the transactions go through multiple processes in the system. All the way through that whole system map, that picture is going to be rewarded and if they will be able to course-correct easily open tracing. Is that specification standard that helps people invest in tracing with low risk and support from open source?

A

Let me show you how but I need your help. Please fill out your mobile phones or any other mobile device that you have and go to donut salon.com, and you will see here a ton of donuts that you can buy hypothetically, unfortunately, and click on as many as you would like to buy. Please help me create a success disaster. Please take out your phones and click on as many Donuts as you would like to buy. You'll see, there's a lot of evading not going on, so the response isn't as fast as it could be.

A

Now, if I look at the architecture for this donuts as a service I have it's a bootstrap project, so I have a web client I have a server and a mixer fryer chopper.

A

Luckily, I have made it with open tracing, though also what I have done is used open tracing baggage tags to tag all types of donuts, so sprinkle, cinnamon, etc. So now, let's see what has been going on? What I'm going to do is use zip gun of well-known vendor for distributed chases and I'm going to find traces here. It might be a little hard to see so I'm gonna zoom in a bit so you'll see majority. The donut fryer seems to be taking a lot of time.

A

Ninety percent- seventy six clicking on this will give me more details which might be a little hard to see, but suffice it to say, fryer is gumming up the system if I click on it I get more details, but I'm, not sure if you guys can see properly anyway, just to check open tracings portability. I am going to go to another vendor like step in this case and see whoa what's happening here. There's a big success disaster. Thank you guys and so I will click on this trace here.

A

What I can see again the same story? The fryer is taking up most of the time. There is more details here and you can see that the cinnamon donut is the gumming up things in the fryer. Now, as you saw it's a lot of information here, so you might be like what's going on, but TLDR is look at this trace. You see the fryer took the most time and you knew that in a second now, let's go back here.

A

So if you're excited about trying out racing seeing if it works for your system, unfortunately I can't tell you anything more. We have only five minutes here so and I'm almost out, but what I can tell you is that we're hosting a distributed tracing salon on Thursday at 2 o'clock in room, a 0 8 where they'll be at racing 101 tutorial where we'll go through an exercise, then we'll go through your specific system and the problems you're trying to solve and they'll be experts available to help you through the process. There'll be good.

A

Mingling people who are performance minded observability minded will be there, but perhaps most importantly, there will be real donuts. So again, please come to a o'clock. Thursday 4:00 distribute racing salon Romeo it thank you I'm Priyanka and please be in touch.