Bestand wählen
Merken

Introducing Chef to an Enterprise and Creating Awesome Chefs

Zitierlink des Filmsegments
Embed Code

Automatisierte Medienanalyse

Beta
Erkannte Entitäten
Sprachtranskript
thank you so will basically be talking about how we are rolled out shaft a little bit of capital 1 some details and will to some best practices that we learned the hard way so that it helps audience over here and how do we sort of take a cloud and Ops and chat journey and its interaction because at among Iraqi on all of us work for their cloud engineering at Capital One and where software engineers so jumping right in the presentation on most of you over here attack on I think I think that cabin 1 is a credit card company indeed we are all we are 1 of the largest in the US with 17 million accounts so if you also might know that kappa 1 it is 1 of the largest banks in the US right and is a leader in banking we have some innovative and cool applications like mobile most of people that have that right but we have app call prior wise which doesn't just look to get credit score but it also you model for a grant score and then we have the few comforts this which can and you can you ought can ask Alexa if I hail give me my account balance and Alexa will let you know account balance and some details on account that our somebody wants to thank you right from some of you the tech community also noise as I that contributors of open source idea common people use video here I've heard about idea you right to high idea is a dashboard arm which is 1 place that you can go configure Our by simple visualization which helps you to visualize your entire workflow but basically your whole pipeline right arm I would encourage people to come and take a look at it is great on dev exchange which is a developer portal the kappa 1 recently rolled out and that gives the ability of fictional developers was our partners to come in and access RTP eyes that includes on 1 of authentication and the cloud custodian that we just rolled out right which helps itself pop the engine ob which does AWS management what policy files and you could take actions upon the resources run in the cloud for more information on these products and you can visit our did sigh or i g excited capital and I am some of you might not know what we are a founded led 20 year old technology company which started added destruction up in the credit industry the company was formed on a premise that all consumer is in the United States should not have the same kind of credit cards how we filled this vision was truly the if we took data
technology data sciences and created information based strategy ossicle IDS which we use it for designing our products if customer needs and the styles verses 81 fit all model we make adjustments to our products and presentations to see the impact in our data would predict business results before a full-scale market we essentially were doing Big Data before the term data came in the picture we are
the largest digital bank in the nation and the prefer channel obviously is mobile we're changing banking for good and adding more humanity ranking for our customers which are looking for the human touch and a smile we have capital 1 cafes the where community can come together know more about capital products or a cup of coffee of founder and
CEO mentioned ultimately the vendors and banking will have the capabilities of a world class software for a company during the last 4 to 5 years will be focused on becoming a world-class suffer shop we mostly belt mouse offers in house worse is out person to become 100 % agile shop worse as utilizing the traditional waterfall we build and automate our software deployments utilizing 3 major pillars I made everything shift laughter and ask for everything the our success in the software development was high but adapting to these principles has helped us with faster and deployment the where the time a concern squashing comes in why cloud in chess right who over here has not gone to the pain of building a server in the dataset aspiration that nobody at almost forces start using cloud day 1 that so yes it has gone through the levels of frustration in building a server in the data center at Chapel on we had a strong and robust pipeline as far as our application development when a couple years ago we realize that now for developers and to make sure that had full capabilities wanted to our stock infrastructure as code and this is where we start on God chaff at to help us out with complete management tool for faster provisioning an on-demand workloads we start utilizing the cloud and this initiated the next-gen infrastructure at kappa 1 we started with a few pilot applications to help us with the reference examples and utilizing best principles offer rolling in public cloud we all of these applications he spun up having nothing cloud at all to running critical production loads in the cloud in a very short amount of time and shaft certainly was a catalyst to do that the now we are focused on further how improve our productivity the move quicker the get things to market faster and continuously improve we build our workloads on public cloud leveraging open source technologies we build using micro-services architecture and RESTful API eyes open source is definitely played a major role in transformation and today we are an open source for shop the we build on open source and we give it back to the open source community the after the success of the pilot applications we accelerated our cloud journey With pardoning the collaborating and empowering are teams at Capital One what are the tools that we utilized was a compare Asian off pipeline starting with a simple app deployment pipeline which altered the press already from the way the they infrastructure provisioning and configuration used to be manually that point right with a simple outflow developer writes code checks ascetical repo the village up picks up the code compiles it knows it run some tests on it and then move it to an artifact repository With infrastructures code similar to act as code infrastructure code also gets stored and every a code repo the bill job picks up the code right knows that run test on it use it to the shaft server or on the confirmation obviously to so 1 of the examples we shared with art application community as well as outside that confirmation rather than the reverse the shuffle thank you capture and you shaft to allow application code the the that that they can created lecture configured application in a consistent matter in a few minutes right so all we wanna make sure that we started creating so chefs right we wanted you knew Shasta around as because there was tons of market I this everybody was excited right is we had this biased off on need from developers to understand have the how can you do that so we decided to start with in-house training versus using traditional chapter heading the I the it's Frank or less but what would we it he's these managed it support that be all also customize their traditional shaft two-day training and make sure we started using our infrastructure as well as pipeline this help that when the developer after the two-day 1 back to his job he basically act as a solution right and they started coding and you start utilizing shaft we also focused on building reusable platform recipes and this helped the shall developers to just take it customize it would give them instructions how to modified by many variables that utilize their five-star recipes that we had already the thought to so how these older strong commit 80 we obviously started using community of practice In the them up space and cloud but it but with
er where horizontal teams will be available for developers to be asking any questions an understanding of that any concerns the technology our product corners start running was of the customers understanding talking to other product corners where they would get more information how better to serve they customers
we have internal pulse site where people come and you could say what features they wanted and specifically they could say this process or this technology the way they use technology is potentially technology Q and will make sure that removed that at Kapil 1 we also
have it are internal stackexchange where the community to share ideas but they could pose questions and had them and other developers since I can respond quickly then obviously open space some with the jerk aim we started with a stand-alone chat server and quickly moved to a tear architectur Due to put our used the are we realize that if we want to use as a convict management tool with scaling and using the full capabilities of the cloud we have to have a high availability Due to not giving our customers any downtime for our applications Chavez working where we have with role that the HA but now working very closely which f to roll out is your time non-time solution it so i'm invite issue now to start getting inside a low the details where he'll share about our our pipeline right and then 3 is gonna continue on getting him the conversation related to the best practices thank you for density followed covering the story of the basics of how you do it on the tool of the cloud of gas 1 and and how we are all you know how others about emissions happening is so that we have the homegrown is it reasonable to the it would safely deployed configurations and if anything went wrong it would stop what was already known and and that he was so OK so that you know for any developer Howard produced by default taking a 4 step so you basically start writing out code on a local workstation and I think that the fastest means to do it's so like any other of good of tool that you use or any other frame of an applications you use of you know you you you just development workstation so in this case we and that is that to use the kid that was provided by the ship and you know how to make as much use of best it's and instead and a shift they put it right at the local workstation and so the 1st thing that is either of us who so yeah so for the very 1st step we get is induced set supermarket so for example and they wanted abilities to be the eyes and business fostered the supermarket and so minutes private or public so we also have a private supermarket implemented within our organization where the developers as deployed and use that as a piece of features that are that you know extends the pollen off interdependency management so as soon as if I use a finanical book that they can use the disuse use bookshelf just would on on the dependency on the local workstation and spin operate instance so all spinning of opinions so it's quite easy using biscuits and so go when the user decides that a z of applicable grounded and spans of of of innocence they can spin it created or destroyed tested again in my divided by and so was that of the very confident that the court is really the goal we and that is and then be without a Amazon E C 2 instance to mimic the environment that they will have in the next it's the production of doing so well you know of 1 that the but also does that accord on E C 2 instances there would be pretty confident that OK it's going good and on the next is as well so after that they can simply be assigned the code and just with get duty for the next 8 so now developping on your logo on you all local workstation is always easy but the next sentence comes when you think of putting icable on a shared shifts over obviously we have a set of steps so of that is shared by my devote all lot of developers and that that brings in more complex against few challenges that around it so all like and that that and obligated company like of banditry guide company their money to buy the things that you should take it off when you're putting on on on a ship on us on of shared server so 1 of those are basically how do you or did you all audio board hard you put this ability in your good idea make sure that the security is that and you know 1 of the most important and it's hard you maintain a good quality standard take across all the Cobol to being of a reduced sets of so we had that only a bigger for application by my but even when we start taking off and faceted as good we think it's important that as well so hard to solve the problem before I started off in a by putting on Nike's RCS several but this 1 that created a cigar development because the CMC had no summer was being used by our system must have developers as well so 100 off then we thought OK let's let's figure out of the way if and that this theory was that basically what it is that you can see in the giddiness and centers nice yes over many of these all of our you know my keys and other credentials of to utilize shifts over and the middle of the was the well and good 5 lines at 100 so that you know and then the rest can make of see out of the cause uneven than the manually built just make of it utilize them and as far their own credentials or bars that on our information to the so of common jobs and just take off and do whatever they or to do but obviously we also introduce user accesses around so that only certain on the final of authorized user can than the jobs and perform the activity and then allowed the amount of all information on axis is that they were supposed to have me the last normal was there it was good and the results b b b also and for the good quality text within the same jobs and up so that although cord across the organization is following the same standard again for and not for the head and my mind let's say you up with the accord and production what an even in Q it the the extended the functionality of you know using of rulers so of ice in the next slides how you can other delivers also just and adult or even at a sort blue 1 of you know any of that information for the the of lower and the upper where we get any man on modification where they can just begin to of you pick to either a prudent rejected based on its unity the all the so here is an example workflow that you can see here of it can see that the this is 1 of the users see on the doing and so so
basically the use of Dinkins as RCS of that because then get is obviously 1 of the most famous tools on the word for this year right well for a CIA pipeline and most of i being so using Jenkins so it was pretty easy to without your job and again so that it was easy to indeed it for the developers of the make happily optimum use of the result it was pretty easy to integrate goes out in the bible and so the the use these out and they don't seem dependencies on this it just made it costly so fun example you can see we had that you and broad specify the of again just big all down the and and all so behind the scenes the up and up you know they don't have to worry about the putting information and I've got be where the system letters and they take care of that the knife used to get of that once again so they can just appointed 1 this up the core depository and descended from the no that was the our scheme foreign this is the screen there are a bluesy the upper west we get an e-mail saying OK of this developer for example in this case the name I'm sure people at the back and see it so there's this of human and so the the upper again just our begin to and see OK this is a promotion that I've got to do I want the would this in production not so they can just execute that it the prof from which and this on on on its validity the now 1 of the main reasons that people and movies these pipelines was we wanted the would visibility and auditability so they're very innovative that's what an on it and the out in all the basically fed all the loss and in vivo close to make sure that the effort ordered the content of better the model you can just look at the dashboard and find out who applauded workable had rule of you know a pool of water ways so you can use any number of who you want on any other that's what will that you're confident that beginning the use Heidi as well but to the gate source level so that we have the likely discussing in only a slight the now the wouldn't all into 1 picture and that this is how the full of books the user pushes in a computes all the development on his local machine and finds out of all the dependencies are satisfied in using supermarket or any of the dependency mental when the user that thing OK my Government is that number put it and thus understand chef so he can look is on job by using his own job you can invoke the the more job and you know just make maybe I got a like a 3rd different manually on uh and published his own goal steps of a market and on the system but at the same time I mean it depends on you but this already doing it so was at Google demands the the user again and you know so the the minute CloudFormation template and all the information from the system so and then the will infrastructure so this is how the lower output by 1 looks like so that's it for myself alone might cilia to the bank of deployed you better information and more information and units should you but hi a you and I like the end of each instead of extremely colleague I bring to this transformation using cloud and develops and especially using shift and how we got a service to local for developers to use it grows organization especially for large company like us we re have thousands of was working on this ship ships right so our analytic over some of the aspects of what we learned on arginine and I think of these as few things of that each of the should keep in his mind before starting his journey on in his journey using ship so it is of perfection comes in practice obviously and as we evolve into doing things are indicative of fashion and we learn from our experiences and from Ireland and so we got a art emission control because convoys abductees and as you and emission control is that the points in the lines of DevOps concluded that had a pointed out last time shift conflict so it's all about that states on the work of learning from your experiences in 6 and the link of recipes speech and every time you know that right so 1 of the key questions and we started with that of a shift is water do that existing automation and what do you do with the existing scripts which I've been working and promising for us for a always make many years right so we cannot just dump all of that stuff and distearate everything and shift and so we need to be that right balance so some of the questions we asked ourselves is how can we leverage existing automation all investment that we get on existing automation and then do we have to really designed cookbooks from scratch on hiding model in which we can leverage part of for automation of was has been there is always a lot of years and leverage the best richer fast right so that's those are the key questions of the as answers when we started the journey and that sea water so this is a simple example when you bought your existing scripts as he's into a chef recipes is a bunch of eggs blocks running the scripts and commands that you usually and when you run your permission to any other operating system but so the problem that this this but you you don't have any control on what's going on in those groups I mean even if you find you're using shift but are you really using the publisher shift which is creating an immutable and idempotent infrastructure right and you also lose that capability off you know avoiding confrontation difficult because these users is none of them are important on you don't have control over what happens to them even if you're in the same this again and you have to do a lot of investment into those steps student aid them to make it look just to make it work which is not what that bank cost especially in this new modern era then you have to deliver it to the so the advantages state like if you wanna use existing strips and this appears as a set you can Kickstarter Archief implementation and you can maybe put your application deployment Automation into production within minutes but what you may lose this use of the data given configuration drift which is a key if of features of the modern infrastructure and he believes that x and lead you to extend the books and create a a a good of community of open-source cookbook writers and yes I mean that the main burning bad design and put into use idempotency and and that
resource which is the which is spoken of from your existing scripts dividends and again and again in it may disappear and dead production system and in human and integration issues as so use users distances I mean this is an example of a disappear being chef resources which are uh which gives you the ability to easily what's in the reason being what it is doing so in many non-technical users on Nantucket person can go through this file answer that OK this file this confirmation of this recipe is installing don't get and its conjugate as a service which has a very simple and very easy to use for it and any notable coming into European maybe years experience which on the a notion of a goal against other countries have what's going on and you can easily use this is again at his own confirmation on top of which is not possible if all of it is very hard to do if it if it if you're just putting your existing scripts into our as shift recipes so we get it out model of each of us best suited for many of beams and the partners doesn't annotations to of get this so the embedding up bank in this is what we did is we separated out and solutions curves was as confrontation scripts and without OK the intuitions clips tends to as is with properly having a guided conditions but the confirmation is must and it should be moved blue shift as templates and cookbook 1st reason for that is the cookbooks say for example bank solution will have its own configuration and everything in the Enterprise may have its own configuration data and we don't want everything enterprise divided on Banca cookbook as so the event occurred 1 cookbook but standard standards and want a good idea to opinions and as and extended and the only way they can easily extend that is by putting the configuration to shift and little resources because that's easily scalable and easily or whatever and easily extensible so and 1 of the other things that I wanted us to quantities ships search so it's not most common uncommon for people to use uh config management to lecture lectures and the service discovery mechanisms on top of search isn't nice features but again it's not the uh giving you service his discovery features or cluster management features that are not so it doesn't have give you consistent users as well I mean it doesn't stop take into account the help of the system so if that if you can just search and the query the shifts a route to get all the service performing of which about of a cluster but you want yearly get busted of a that that someone is has been of predict the traffic or you end up confering that so even if in in in its and hence the state into that lasted and it starts taking traffic which again is that right so shift search is not ideal for service discovery that's what we learnt or Argentina using shift and it proper service discovery to when they considered to be was a component of that service discovery goes job range of does config management and automation for you and making the separation between these duties of a given our paper was infrastructure and adding up so as to deliberately of what you think is needed in order for us to operate recipes perfectly each and every day so 1 thing that he told me is this the the well-being and a mobile learning at the speed of the velocity and the requirements and infrastructures changing the fast rate so he said if you want to have a little consistent is seduction of operational notion of cookbooks there is a need for automated bill and this pipeline the dot which you end up performing than it does for the features that you ordered according to a use of 2 years back so the importance of it blamed for automated testing is manner testing is a plausible and it is effective but it is limited and mean as you as you add more releases your permission it's not really scalable practical to perform an interesting you do not want to do then managed this affiliate for just a new feature that you added this cycle rate you want all of those staying all features to be tested automatic and you wonder just as the new feature that you create least recently so that's the necessity of having an automated the Bible and it ensures that the assistance and contributions are safe and their free and uh most often you would be dead there is encoded them out very early in your life cycle or development cycle so this is an example and 1 bed lay in that 1 of our teams of bird of it shows the embeddings of having a biplane right from the core commit under look Department late and it now and the quote for that matter regardless of whether it's application infrastructure most of the same sort of status and the same set of against uh in the pipeline and that the and shows that their and their infrastructure and application of blue consistently and faster so uh just a quick summary of what we discussed a very late so building stronger communities is a must for anyone to desert plans to transform into it develops shop or develops so so practices and a standard look for a visit with best practices and shows that you're up complaining the good organizational policies at the same time I leveraging the benefits of the Robson and the new technologies that we using disease that's what you do what's gives you feedback real-time and that is the most powerful european spinning in order to get feedback very early in the development cycle and start beginning to to excuse me so automated testing is a must as is it earlier and you have to create biplanes replications infrastructure just like I did them both the same because their query infrastructure scored an application is also called and once you create and trade them together and further through the same pipeline that you use for your applications you and ensure that you have a robust consistent and immutable perception can finally locity and become a message of so far so it Gadolin of know pushing the limits of innovation and using technologies like loud DevOps open source and in contributing back to open-source really out the other thing too is sort of but we don't find in NBA the using the Commission and predict open-source products of which are of the this is a job and forest and saying that the just things the thank
you you and is curious we can think about the shit out of me to say that provides the visibility you have you so what we have not explored the a tool just yet right so so can answer will hopefully the looking at it on I could talk a little buckshot delivery from we have our like we started with the DevOps principle so we had a strong and sorry to echo at sensing something is messed up so we have our very strong pipeline minority and so we decided to continue with Jenkins and that's doing the job right what's the lovely probably picks up a little bit and has more features will pulsars in looking at this point is using Jenkins have you published information about how you've actually done the automation of life we are and that's 1 the things when we were preparing the presentation came true that will probably will be publishing that because that's for useful information and will therefore mu on working on open source in I think you even on applause Thank you
Offene Menge
Retrievalsprache
Bit
Gruppenoperation
Datenmanagement
Geräusch
Kombinatorische Gruppentheorie
Unternehmensmodell
Kappa-Koeffizient
Videokonferenz
Unternehmensarchitektur
Datenmanagement
Endlicher Graph
Visualisierung
Softwareentwickler
Maschinelles Sehen
Konfigurationsraum
Streuungsdiagramm
App <Programm>
Nichtlinearer Operator
Open Source
Systemaufruf
Elektronische Publikation
Biprodukt
Chipkarte
Summengleichung
Software
Rechter Winkel
Chatten <Kommunikation>
Authentifikation
Information
Software Engineering
Streuungsdiagramm
Resultante
Maschinenschreiben
Güte der Anpassung
Rotationsfläche
Biprodukt
Kombinatorische Gruppentheorie
Term
Unternehmensmodell
Softwaretest
Prognoseverfahren
Eindringerkennung
Strategisches Spiel
Information
Fitnessfunktion
Konfiguration <Informatik>
Punkt
Wellenpaket
Klasse <Mathematik>
Gebäude <Mathematik>
Kartesische Koordinaten
Transformation <Mathematik>
Systemplattform
Raum-Zeit
Code
Kappa-Koeffizient
Übergang
Rechenzentrum
Variable
Datenmanagement
Reverse Engineering
Prozess <Informatik>
Software
Geometrische Frustration
Code
Softwareentwickler
Konfigurationsraum
Hilfesystem
Verschiebungsoperator
Streuungsdiagramm
Softwaretest
App <Programm>
Vervollständigung <Mathematik>
Dokumentenserver
REST <Informatik>
Open Source
Gebäude <Mathematik>
Systemplattform
Biprodukt
Beanspruchung
Software
Computerschach
Forcing
Rechter Winkel
Last
Grundsätze ordnungsmäßiger Datenverarbeitung
Server
Information
Repository <Informatik>
Unternehmensarchitektur
Streuungsdiagramm
Resultante
Unterring
Umsetzung <Informatik>
Prozess <Physik>
Gemeinsamer Speicher
Kartesische Koordinaten
Komplex <Algebra>
Raum-Zeit
Eins
Puls <Technik>
Datenmanagement
Prozess <Informatik>
Code
Statistische Analyse
Elektronischer Programmführer
Default
Gerade
Metropolitan area network
Verschiebungsoperator
Zentrische Streckung
Lineares Funktional
Computersicherheit
Verhandlungs-Informationssystem
Güte der Anpassung
Hochverfügbarkeit
Stellenring
Biprodukt
Dichte <Physik>
Rechenschieber
Menge
Rechter Winkel
Ablöseblase
Server
Information
Programmierumgebung
Schlüsselverwaltung
Instantiierung
Standardabweichung
Server
Web Site
Rahmenproblem
Selbst organisierendes System
Physikalische Theorie
Whiteboard
Code
Arbeitsplatzcomputer
Biprodukt
Softwareentwickler
Konfigurationsraum
Schreib-Lese-Kopf
Transformation <Mathematik>
Schlussregel
Physikalisches System
Quick-Sort
Streuungsdiagramm
Einfügungsdämpfung
Idempotent
t-Test
Datenmanagement
Unternehmensmodell
Hinterlegungsverfahren <Kryptologie>
Softwaretest
Dämpfung
Perfekte Gruppe
Prognoseverfahren
Standardabweichung
Datenreplikation
Skript <Programm>
Notepad-Computer
Gerade
Verschiebungsoperator
Softwaretest
Kraftfahrzeugmechatroniker
Oval
Biprodukt
Bitrate
Ereignishorizont
Gruppenoperation
Menge
Einheit <Mathematik>
Rechter Winkel
Konditionszahl
Server
Ordnung <Mathematik>
Geschwindigkeit
Selbst organisierendes System
Wasserdampftafel
Automatische Handlungsplanung
Abgeschlossene Menge
Demoszene <Programmierung>
Virtuelle Maschine
Spannweite <Stochastik>
Netzbetriebssystem
Skript <Programm>
Inhalt <Mathematik>
Abstand
Ideal <Mathematik>
Konfigurationsraum
Demo <Programm>
Videospiel
Open Source
Schlussregel
Binder <Informatik>
Elektronische Publikation
Summengleichung
Gamecontroller
Wort <Informatik>
Eigentliche Abbildung
Geschwindigkeit
Streuungsdiagramm
Resultante
Punkt
Gruppenkeim
Versionsverwaltung
Kartesische Koordinaten
Computerunterstütztes Verfahren
Übergang
Einheit <Mathematik>
Datenmanagement
Prozess <Informatik>
Kurvenanpassung
Funktion <Mathematik>
Nichtlinearer Operator
Topologische Einbettung
Template
Abfrage
Nummerung
Programmierumgebung
p-Block
Verknüpfungsglied
Funktion <Mathematik>
Information
Versionsverwaltung
Message-Passing
Aggregatzustand
Standardabweichung
Server
Implementierung
Zahlenbereich
Gebäude <Mathematik>
Sprachsynthese
Transformation <Mathematik>
Systemprogrammierung
Physikalisches System
Selbst organisierendes System
Unternehmensmodell
Inverser Limes
Zusammenhängender Graph
Softwareentwickler
Hilfesystem
Implementierung
Touchscreen
Streuungsdiagramm
Trennungsaxiom
Fehlermeldung
Wald <Graphentheorie>
Idempotent
Konfigurationsraum
Validität
Routing
Physikalisches System
Hybridrechner
Quick-Sort
Integral
Modallogik
Touchscreen
Mereologie
Dreiecksfreier Graph
Speicherabzug
Innerer Automorphismus
Unternehmensarchitektur
Videospiel
Bit
Punkt
Rechter Winkel
Prozess <Informatik>
Open Source
Information
Kombinatorische Gruppentheorie

Metadaten

Formale Metadaten

Titel Introducing Chef to an Enterprise and Creating Awesome Chefs
Serientitel ChefConf 2016
Autor Avireni, Surya T.
Gupta, Ishu
Ravji, Ali
Lizenz CC-Namensnennung - Weitergabe unter gleichen Bedingungen 3.0 Unported:
Sie dürfen das Werk bzw. den Inhalt zu jedem legalen und nicht-kommerziellen 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 und das Werk bzw. diesen Inhalt auch in veränderter Form nur unter den Bedingungen dieser Lizenz weitergeben.
DOI 10.5446/34637
Herausgeber Confreaks, LLC
Erscheinungsjahr 2016
Sprache Englisch

Inhaltliche Metadaten

Fachgebiet Informatik
Abstract At Capital One, we want our cloud-enabled infrastructure to be an incubator for innovation and an accelerator for bringing more capabilities to our customers. We embraced the principles of Automation, Agile, DevOps, DevOpsSecurity, and Open Source with a robust automation framework to reach our goals. Chef combines innovation, speed, collaboration, and safety all into one DevOps platform. We introduced Chef to our DevOps engineers and quickly built a strong user community through sharing code and discussion forums like office hours and an internal Stack Exchange. Our Chefs didn't need to keep a personal knife because our Jenkins did all the work. We built a flexible Jenkins pipeline to deliver cookbook-enabled integration with automated application builds and provisioning. Implementing Chef Analytics provided more insight into the actions of the nodes and fed all of this data into Splunk for better visualization. A highly available Chef server and a private Supermarket provided our DevOps engineers with everything needed to manage their infrastructure and share their automation. This enabled fast and flexible IT as well as continuous delivery of applications and infrastructure. In this talk, we will share some details about our journey from sous chefs to master chefs. We hope you can leverage our experience on your own master chef journey.

Ähnliche Filme

Loading...
Feedback