Advancing the Future of CI/CD Together
This is a modal window.
Das Video konnte nicht geladen werden, da entweder ein Server- oder Netzwerkfehler auftrat oder das Format nicht unterstützt wird.
Formale Metadaten
| Titel |
| |
| Untertitel |
| |
| Serientitel | ||
| Anzahl der Teile | 490 | |
| Autor | ||
| Lizenz | CC-Namensnennung 2.0 Belgien: Sie dürfen das Werk bzw. den Inhalt zu jedem legalen Zweck nutzen, verändern und in unveränderter oder veränderter Form vervielfältigen, verbreiten und öffentlich zugänglich machen, sofern Sie den Namen des Autors/Rechteinhabers in der von ihm festgelegten Weise nennen. | |
| Identifikatoren | 10.5446/47474 (DOI) | |
| Herausgeber | ||
| Erscheinungsjahr | ||
| Sprache |
Inhaltliche Metadaten
| Fachgebiet | ||
| Genre | ||
| Abstract |
|
FOSDEM 2020329 / 490
4
7
9
10
14
15
16
25
26
29
31
33
34
35
37
40
41
42
43
45
46
47
50
51
52
53
54
58
60
64
65
66
67
70
71
72
74
75
76
77
78
82
83
84
86
89
90
93
94
95
96
98
100
101
105
106
109
110
116
118
123
124
130
135
137
141
142
144
146
151
154
157
159
164
166
167
169
172
174
178
182
184
185
186
187
189
190
191
192
193
194
195
200
202
203
204
205
206
207
208
211
212
214
218
222
225
228
230
232
233
235
236
240
242
244
249
250
251
253
254
258
261
262
266
267
268
271
273
274
275
278
280
281
282
283
284
285
286
288
289
290
291
293
295
296
297
298
301
302
303
305
306
307
310
311
315
317
318
319
328
333
350
353
354
356
359
360
361
370
372
373
374
375
379
380
381
383
385
386
387
388
391
393
394
395
397
398
399
401
409
410
411
414
420
421
422
423
424
425
427
429
430
434
438
439
444
449
450
454
457
458
459
460
461
464
465
466
468
469
470
471
472
480
484
486
487
489
490
00:00
Stetige FunktionRichtungOpen SourceGüte der AnpassungXML
00:16
WhiteboardPunktwolkeSpider <Programm>Prinzip der gleichmäßigen BeschränktheitWhiteboardOpen SourceTwitter <Softwareplattform>SystemaufrufCodeAnalytische FortsetzungPunktwolkeComputeranimation
00:37
AggregatzustandRegulärer GraphComputeranimation
01:14
ProgrammierumgebungSoftwareQuick-SortAutorisierungInformationMapping <Computergraphik>AggregatzustandLeistung <Physik>QuadratzahlMultiplikationsoperatorSystemaufrufVerkehrsinformationComputeranimation
02:29
MathematikDienst <Informatik>ProgrammierparadigmaKartesische KoordinatenComputersicherheitVerschiebungsoperatorMomentenproblemDifferenteDienst <Informatik>BitMechanismus-Design-TheorieMathematikQuick-SortGruppenoperationPortscannerComputeranimation
03:51
VerteilungsfunktionPunktwolkeOpen SourceMinkowski-MetrikProjektive EbeneNichtlinearer OperatorOpen SourceComputerspielAnalytische FortsetzungÜberlagerung <Mathematik>Computeranimation
04:36
Kollaboration <Informatik>Virtuelle AdresseAnalytische FortsetzungGenerator <Informatik>Kollaboration <Informatik>
04:51
WhiteboardProjektive EbeneVollständiger VerbandWhiteboardComputeranimation
05:33
PROMWeb SiteComputeranimation
06:02
Analytische FortsetzungUltraviolett-PhotoelektronenspektroskopieSchnittmengeMeta-TagVersionsverwaltungVollständiger VerbandMinkowski-MetrikWort <Informatik>Open SourceFlächeninhaltTermProjektive EbenePunktwolkeKategorie <Mathematik>PunktKonfiguration <Informatik>Computeranimation
07:11
TermAnalytische FortsetzungProjektive EbeneSchnittmengeWort <Informatik>FlächeninhaltHypermediaFamilie <Mathematik>Computeranimation
07:46
Lemma <Logik>FlächeninhaltComputeranimation
07:59
UmwandlungsenthalpieProjektive EbeneMereologieARM <Computerarchitektur>Minkowski-MetrikComputeranimation
08:29
ComputerspielProgrammierparadigmaDifferenteQuick-SortMereologieBildschirmmaskeFormale SpracheDienst <Informatik>Rechter WinkelComputeranimation
09:49
SoftwareentwicklerProjektive EbeneQuick-SortComputeranimation
10:16
Interface <Schaltung>p-BlockGebäude <Mathematik>Ganze Funktionp-BlockShape <Informatik>Gebäude <Mathematik>MetadatenGanze FunktionAggregatzustandProzess <Informatik>GruppenoperationComputeranimation
11:08
VersionsverwaltungGemeinsamer SpeicherGruppenoperationProzess <Informatik>Quick-SortMailing-ListeDifferenteArithmetische FolgeProjektive EbeneÄquivalenzklasseComputeranimation
12:14
Web logFlächeninhaltBitAlgorithmische LerntheorieGruppenoperationMAPVirtuelle MaschineNichtlinearer OperatorComputeranimation
12:52
Maschinelles LernenMaßerweiterungProgrammierumgebungAlgorithmische LerntheorieVirtuelle MaschineMaßerweiterungGruppenoperationProdukt <Mathematik>Notebook-ComputerEndliche ModelltheorieCodeComputersicherheitMultiplikationsoperatorKlasse <Mathematik>ProgrammierumgebungVollständiger VerbandFlächeninhaltComputeranimation
13:42
Element <Gruppentheorie>MenütechnikRichtungComputeranimation
14:08
Gebäude <Mathematik>Web SiteOpen SourceDifferenteTwitter <Softwareplattform>Exogene VariableApp <Programm>Computeranimation
15:25
FacebookPunktwolkeOpen Source
Transkript: Englisch(automatisch erzeugt)
00:05
OK, well, good morning, good afternoon, everybody. I'm here to talk about how we can advance the future of CI CD together and under the umbrella of the CD Foundation. So just first up, I'm director of open source at CloudBees.
00:21
I'm also on the governing board of the Continuous Delivery Foundation. I'm the co-instigator of the great Brexit pub crawl. Was anyone here at that on Friday? And you can find me on Twitter as Tracy Miranda. So I'm here to talk about continuous delivery, which
00:42
we saw in a poll earlier. That's what many people see as the CD of CI CD. And it's essentially an engineering practice where teams commit regular amounts of code, but they always want to keep it in a state
01:01
where it's ready to ship. And so continuous delivery is something that, as a concept, has been around for 10 years or so. And we have a lot of research in this day and age about the best way to deliver software. So who's read the Accelerate book or come across that?
01:24
Yeah, a few hands. But I highly recommend that book as the modern text for how you and your team should be delivering software in this day and age. And that's accompanied by the state of DevOps reports, which come out sort of every year.
01:41
And they give information about ways you can measure your software throughput, but also really good suggestions about how you would improve that in your team. And yeah, one clue. It's more about the team environment than the individuals and their special powers. So I love that quote from Nicole Forsgren,
02:01
one of the authors of the book, who just talks about how the people, the teams that are good at it just keep getting better and better. And we have teams that can kind of really struggle if they don't sort of get the right things together. So continuous delivery has been around for a while. We have a book which maps it all out.
02:22
But at the same time, we find that adoption is really, really low for this practice. And we go, well, why is that? And there's three big challenges that I think people are facing at the moment. And the first is this huge shift in the industry, which is the rise of microservices.
02:40
So who's doing things with microservices today? I'd say about half the room. So that presents some interesting challenges. You no longer have your monolith application. What does it mean? When are you releasing different bits of the services that make up, what even is your application?
03:01
How do you define it? And then you've got cloud-native technologies, things like Kubernetes. Who's using Kubernetes at the moment? Yeah, quite a few more. That introduces a whole kind of distributed paradigm, which has different challenges to contend with. On top of that, you look at the tool landscape today,
03:22
and it's pretty fragmented. You have all sorts of different CI and CD tools. They don't necessarily work together, and it's often left to you as the DevOps folks to work out how to integrate, I don't know, a security scan with your bills, with your delivery mechanism. And then finally, of course, change is hard.
03:42
So we're all human. We all have teams. We all like how we do certain things. And just having to evolve that as a group can be pretty challenging. So that was some of the motivation for why the Continuous Delivery Foundation was formed. It's less than a year old.
04:01
So this is the first FOSDEM it's really been alive at. And as an open-source foundation, it started as well around four founding projects. And I'll talk more about those in a minute. But it was pretty significant just because we were kind of trying
04:20
to bring together the whole space. So we call it Continuous Delivery Foundation, but it does cover CI, continuous deployment, continuous delivery, and DevOps. And we're looking at different ways we can tackle problems in this space. So essentially, we see ourselves as a neutral home
04:40
for the next generation of continuous delivery collaboration. And that's the key benefit of how we're gonna come together to solve industry-wide problems. So one of the ways we've been thinking about what can we do as this foundation after we launched was to get a lot of the board members together.
05:02
And you might recognize some of the project leads, Kosuke Kawaguchi's in there, of Jenkins and Andy Glover of Spinnaker. And then we've also had Jez Humble, who wrote the book on continuous delivery. And he sat down with us to share his experience as well as think about how can we go about
05:22
solving these problems. And the thing I love about this picture is that everybody's wearing shades because the future of CI-CD is so bright. So the outcome of that meeting was actually to come up with nine goals for the foundation. And so there's a lot there,
05:42
but what I'm gonna do is pick out about three or four of those and tell you where we've gotten to in our journey and how you can get involved if it's something interesting to you. For the full details of all the different goals and how we are trying to achieve them, you can check out the website at cd.foundation.
06:03
So the first one is driving continuous delivery adoption, helping people in their journeys. When you come to the landscape, the first thing, like the CNCF, which is the Cloud Native Compute Foundation, they have this massive landscape
06:20
which shows all the different tools and how they get qualified. We also decided a very first step is to start to try to categorize CI-CD tools, which turns out is really, really hard because things fit in multiple categories. But we do have a first version of the landscape, which is open source and people can make contributions.
06:41
We're hoping to get all the different tools in there and to mark them in the different categories, so the main thing of what they do. So I'd invite folks to check that out. Definitely has mistakes, so you're welcome to submit a pull request, but it's a version 0.1 and we're hoping to build this up to give, just get people new into the space,
07:02
start to have an idea of what the different categories are, what they mean and what are the options in the space. We also, just in terms of spreading the word about continuous delivery, we have a set of meetups. So these were actually Jenkins area meetups.
07:23
So Jenkins is one of the founding projects and these meetups were contributed to the CD Foundation and the CD Foundation is now kind of driving those forward. Meetup owners are invited to widen the scope, so if you want to talk about more than Jenkins, you can do that and we've got quite a few all around the world
07:41
and maybe some of you have already been there, maybe some of you run them, but either way I encourage you to either join a local meetup or look to start one yourself so you can sort of work together with other people to sort of see how people are approaching things.
08:00
Okay, then one of the other missions is to just focus on the projects that are part of the foundation. So there's a lot of CI CD projects out in the world. We launched with four specific projects, so Jenkins, Spinnaker, Tecton and Jenkins X.
08:21
So these are founding projects and the interesting thing I find about the different projects we have in the space is that they really represent different parts of adoption in the industry. So you take something like Jenkins, okay, who's heard of Jenkins?
08:41
Yeah, I think some of you are just being lazy. So most people, I say Jenkins is pretty much late majority if you're doing some form of CI CD, you've come across it, you've used it and it's only laggards who perhaps aren't employing CI CD who aren't using it. Spinnaker, probably early majority,
09:01
lots of people using Spinnaker. Well, how many people in this room are using Spinnaker? I've heard of Spinnaker. Okay, maybe. Yeah, just a few people. And then over here on the right, Jenkins X and Tecton, for those of you who were in the talk earlier, we talked about Tecton and there'll be some talks later about Jenkins X.
09:21
Those are coming in much more on the kind of innovators and early adopters, folks who are just getting to grips with microservices and Kubernetes and looking for tools that can deal with the different kind of tech paradigm. So tools in different languages, tools with different challenges. So as a foundation, we're sort of saying,
09:41
okay, what does each tool need and how can we help them grow and be useful to the end users who want to use them? And some of the things we do is just measuring how many contributions we have. That was a little infographic I did, just trying to compare how many people are contributing to the different projects compared to, let's say, something like Kubernetes,
10:02
which has really taken off. And just saying, like, you know, Jenkins, for basically a 15-year-old project, is doing pretty well and quite healthy, and all the other projects are sort of growing steadily. So that's something we'll look at sort of year on year. But a key thing is we do want to
10:20
just foster some tool interoperability. So how do we make sure all these tools stay working together? So we want to think about, you know, standardizing, building blocks into shared APIs, building this whole ecosystem where things can be plug and play, so it's not like you have to go and figure out how they all work together for yourself.
10:42
And then just generally, we see that will improve the state of delivery for the entire industry if people are not left struggling to themselves and working out what's gone wrong when they plug tools together. So we have some ideas for common APIs and establishing, let's say, common metadata. You know, how do you define a release?
11:00
It should be standardized, whether it's from Spinnaker or whether it's from Jenkins X. So we're looking at ways we can do that. So very recently, we spun up a working group called the SIG Interoperability Group. And where's Fattie? Oh, there we are. So Fattie is one of our chairs
11:20
who's doing an incredible job just getting people together. And I wanna highlight this full request, which is in progress. And it's one of the first things which is actually saying, look, there's all sorts of different tools out there. And even just with the vocabulary, what do they refer to as a pipeline?
11:41
What does it mean for that tool? So can we, first of all, define them and then maybe start to translate so that if you're talking about a pipeline in one tool, is that equivalent? Is the step the same thing? So before we can integrate, we need to just sort of say what is it? What are the different tools talking about?
12:01
So again, this is all in GitHub. I encourage you, if you have a project, to go and add it to the list and to just contribute to that sort of first step in getting to the shared vocabulary. Okay, and final bit in this kind of whirlwind tour
12:21
is the goal of expanding into emerging tech areas. So machine learning is something that's emerging in quite a few areas. And we have a specific area, so machine learning ops is the intersection of DevOps and machine learning.
12:41
So we've also got a special interest group that started up around that. I believe we've got Cara De La Marque is one of the very active members in there. And so one of the things I'd like to highlight from that group is that group is working on a very initial roadmap.
13:01
So what does it mean for DevOps and machine learning? So they talk about in the roadmap, if I pick out two highlights, what is MLOps? One, it's the extension of the DevOps methodology to include machine learning. So making sure models and data and code are all first-class citizens.
13:22
And then things like what is it not? It's not about putting Jupyter notebooks into production environments, but helping people find better ways to do things. So I highly encourage you to go check that out. And again, if that's of interest, the group is open, join the meetings, talk to Cara, and yes, it's pretty exciting.
13:43
Okay, I think that's all I have time for today. So I haven't even touched on some of the other things we're doing around diversity and security, both very important areas. But I hope you get the general idea that it's just people coming together, trying to drive things in the right direction
14:03
for the good of the full industry, and not just one specific tool or one specific vendor. And yeah, please check out our website and come try and join us. And we also have a Twitter account where you can find out what's happening with the foundation.
14:20
Thank you very much. Any question? One question. Okay, go ahead. I've seen an overlap with what Kubernetes SIGs are doing. Are you collaborating with SIG apps on the app definition?
14:42
Yes, so we do want to work not just with CNCF, but all the different bodies. Oh, okay. So the question was, we're seeing a lot of maybe overlap with what CDF is doing, and other communities like CNCF and the SIG app delivery. So my response is yes,
15:00
we're looking to work with a lot of different open source communities out there. CNCF is one, and actually in our interoperability SIG, which has just recently started, we did have a member of the SIG app delivery join. So we hope that we continue talking and just find ways to work together without necessarily all working on the same things
15:21
in different ways. Yeah.