Merken

Introduction to Baserock

Zitierlink des Filmsegments
Embed Code

Automatisierte Medienanalyse

Beta
Erkannte Entitäten
Sprachtranskript
so please join me in welcoming so that however the is that to allow to quiet that's OK billion of thanks for coming to this talk I'm going to be talking about integrating software and collections of software that work together returns easy sounds like a solved problem but actually is isn't so many people have the people do that they're um working distributions for creating suffer from that it's systems and the people a few so
do find it easy because of the pain points this effect so that the basic project is developing tools to try and make this process easier uh the goal isn't to replace traditional distributions the goal is to develop a kind of a research project to develop a set of parts which work together but also work independently so people can adopt any of the parts that they see people for them although the tooling is written in Python and it's all free of legacy for the project started about 4 years ago so not completely free of legacy more free of legacy than most of the existing things in this area said the project started with this problem build a working day Linux operating system straight from the source code and how many if I asked how many lines of Python code using you take me to do that any guesses my guess is well I'll give a couple of hints so the predictive dealing with source codes we have a solution for lot which is a service which mirrors every form of version control every popular form of version control and there's tarballs all into get repositories 1 server so the build tool doesn't have to deal with downloading cobbles from London places where thing else they can consider that everything's in gets most things are in get now but importing from material suppression and whatever else they get a consistent interface and then also all the build instructions that you need spelled out In this consistent Yamal formats so this is an example of a simple build instruction from the tells this is the instruction of a Python while the Python interpreter so it says is the standard commands tools but override the configured commands run something when it finishes to create a some so we have a a reference distribution which describes how to build the whole system inform like this there's another slightly more complicated Yamal document which then says what raft to build an artificial together that's it and so With these parts as the other build tool which produces a work operating system with the pirates 2 thousand lines of Python which is an order of magnitude simpler than anything else you find in this area think and that's because retaining the approach that writing a build system should be easy enough that this squirrel monkey could do it if we solve the problems around all the problems in the area than the bill to itself becomes trivial which is good is writing to build tools got a thankless task nobody really enjoys doing it so if we remove all of the the problems around it and it becomes trivial at least 1 not trivial but very easy lines of code is a bit of a horrible metric and I don't may assigned to much meaning to it the that the telling question is kind of prototype is called YTD and we have an older build tool is welcome off which is a lot bigger I'll go through the bits of space rock so what you need to actually build such a system this is the the items that you need really I'll go through each of those in the more detailed so in um this is called mirroring service that's a server appliance called from there we have 1 running which I should be able
to show you in a browser that is when we have why get don't they struck org that contains lots and
lots of good repositories quite boring but it's good to have a consistent interface and and
this is way you submit a patch against of Republika lorries and if there is more things you can also set up your own instance of this for the
actual mirroring to a lot of this is a simple script called lorry which takes the adjacent followers describes where to get source code from and pursuant to get so that's just curve mirroring you then need a way of describing build instructions on the go into that in more detail later on thing is 1 the most interesting part of the project Summer touch on that you then a build tool traps really patterns and Ronald need a language for creating build instructions and then you need some actual build instructions so separated we've defined the syntax for describing how to build stuff and then In order for the tool to be useful we have a set of definitions that you can use to build a system as well but you don't have to use those definitions so this is is quite hard to visualize build systems and build tools so I apologize for the fact that a lot of my slides screenshots of a terminal then I also have a few diagrams but this is the list of package groupings recalled there like ladies in that bike and you can see if the text is big enough some fairly standard packages GTK experienced Python libraries OpenStack is in there you can actually use space tools to deploy in OpenStack you know instance which is composed of a thing and going back to building a working in a Linux system we released every so often we have
released when the reference systems
the build reference system which you can download from here also just to
show that it does in fact work this is maybe loading
the free energy in QMU so this is a base rock reference system and it leads to a Bosch prompt and the reactants a Linux system FIL entirely 1 cool thing about stuff 1st but with base truck is every image contains metadata that shows you exactly what repos MRF everything was built from so this is slashed by short directory is big enough the like is it's QMU company make any bigger without using a serial console so I apologize if you can't read it that's there's a bunch of metadata files 1 for each component in the system and then fired upon them briefly it contains metadata apart this is the set of components and it was built from it was built with these build instructions this was the environment and these are the versions of the dependencies and at the bottom it shows you the URL of the repo it was built from and exactly what shall 1 it was built so if you've got the source code marriage In the show the server showed you then you can go from any system has been built and you can look at exactly what committee of what get repository everything we built from so this no system broken can't actually work out what I'm running that problem goes away and so we have to build
instructions therein a repository and then get basic dog McCullum definitions then we have a tool in fact utils which you can use to build them so source code from gate goes in the build tool just once a sequence of shell commands in the right order which like as a should be easy and produces a binary and then we have an artifact cash which just holds toggles of binary so this 2 tools morph is the old 1 which has a lot of features some of which it doesn't actually need it turns out and there hasn't quite cool things you can as a distributed plugin so you can set up multiple morph workers and have shared builds at the component level so if you've heard of DCC which distributes at the level of the so far North can distribute to the level of the actual component so you could have different packages compiling and different systems and why is more of a proof of concept that shows that you can make a radically simple filter they're both as well the available on doctor-patient dog so after you build something is so much use having a terrible really so you then need a tool to deploy it and deployment is a bit more messy and building thing building is quite a well-defined problem as a building I should be saying building and integration because there's more to it than just running compilers but the output is a binary when she tried to run the binary need to do some extra work to deploy it so for example if you wanted deployed to OpenStack you need to create a disk image upload that as an image to you OpenStack bonds and then if you want to deploy it to dock and then you need to import into darker is it often if you want to deploy it to the real hardware than you may have to put on a nasty God wait 5 minutes state the SD card up in machine so deployment is a bit more messy and there is to bring him a structure do that at the moment but actually I'd like to get rid of it completely much how many people had no runcible it's suppose great much like a really like to replace our deployment functionality with financeable module so we have to think about and the shock anymore because I think and simple souls love problems really neatly so we have deployment the last piece of the puzzle is because you you don't build things more than once and because of the way by structure at the inputs of everything it builds and builds everything in an isolated staging area like an isolated treats you can be sure of that what if you run the same bill twice you get the same thing out not always the same the scholarly are working on that but you get after the artifact which works the same each time so you can cash basically by hashing all of the input signal of the dependencies coming up with an identity and then saying right this is what I built I'll refer to this hatch and if something's only bills then it's already cached and you don't need to build it again so we have a simple cache server which you can use for storing now there are a couple of other bits the we have a continuous building which is really just a shell script which one's more filter 11 over again it's it's not that interesting come and talk about we have attitudes about sandboxing builds we recently spun out the code to do sandboxing into a simple python library called sandbox so it has 1 the API basically as 1 function call and a couple of others 1 main
function call which runs a commands like process stopped open but you can specify a couple of things you do don't want to share all isolates so you can say but in a new isolated map space so we can't see the from the system all put in a new network namespaces can't connect to the Internet and among these extra directories from the host or makes a bitch read-only it doesn't implement of functionality itself because there's lots of tools that already do it but they have different strengths and weaknesses for example lots of the containerisation tools like Dr. and system and born rocket need to be run as
well you can use a much simpler to of Linux User Chile and run use most of those of the next specific so also has a cheery back and which will run on any Posix but doesn't support most of the somewhat and capabilities because in a treaty can't for using just positive API you can't say and of a new man namespace because they're the next specific feature so the trip back and is fairly incapable but allows you to degrade some watching capabilities if you want so this is the tragic again filled in with the names of some component French so I that the most to the part that interest me the most about the structure is the the definitions language which we refer to as declarative build instructions or declarative definitions and the idea is to turn build instructions into data at the moment that code there's lots of build instructions in the world I mean that in build instructions for 10 thousand or 100 thousand packages but it's all covered it really hard to reason about it unless you understand all 7 build systems that have been developed over the years whereas declarative build instructions you wanna treat the build instructions simple sequences of commands so they can be treated a lot more like dates and this we discourage kind of hot implementing features in shell scripts in the build instructions the this no logic for the build tool mixed in Geneva could build rate which is a tool toolbar and largely make for building and Syst instances code build really great that nobody really understands how to how the core of it works anymore because all of the instructions are in make and tied up with the bill definitions themselves and so why that works it's quite difficult to make changes to anymore and finally really don't like shell scripts so I'd like to minimize and resolve scripts in the world but much of that everything is so that data and Python script so what we've done is defined this language young languages define a few years ago monocular rationalize and turn into something formal and useful outside by we have a I defined the schema of the current data model a try to make a nice graph and instead a kind of this graph which shows you the entities we have a the minutes so we have of the command sequence that's the fundamental unit of building something here a sequence of commands for example configure make make install and then there's something called the chunk which is kind of like a package we have these grouping called strata and systems which I think in the future will do away with the just have 1 sort of component that contains all the components really I think the main problem in the network is coming up with a word which means component that contains the components without having it be really long all really weird naming things is 1 of the hardest problems in computer science so the moment we have this data model which is still fairly simple the final entity the cluster which represents a cluster of systems so when you deploy something with basic tools you deploy a cluster it is only 1 and so we have our reference systems repository contains a set of chunks for things like Python gtk cute and different Python libraries it contains strata which integrates those into logical groupings so for example is a cute stratum which contains the various executes that you need to use it systems rich have a specific purpose so for example the OpenStack service system contains a bunch of different things its purpose is to deploy in OpenStack system that you can then host at the ends in but there's also a build system which has built tools in such things and it's easy to define your own ones I'll show you a show
this year actually I was going to show you why BD starting to build something they were finished because then will take hours and I'll probably have time but this is the
data definitions of reference definitions repository and systems structural recommended it we define in the
build system and it contains a simple list of the strata that you want so for example call Python libraries and the BSP which contains Linux and a boot loader different Python libraries unstable colonists in its and things and a photo it's monitored photo why BT to build up we all unintelligible how far object so much from connected the international monetary fund the as the loading things from disk in come back to that another interesting
thing we can do with once the
definitions are considered data is this a lot of existing data analysis tools which you can use to look at them so I made this this is why we do we
actually building something so calculated as identity for each component involved in the bills and pretty soon i'll get to the point of running some running configure further in utils probably 0 yeah so this is 1 of a short build tool actually looks like this is running command and this will take about 4 hours will give a systematic decide which I would show you the to go much browsing the
definitions I found also in Python library called RDF live web so RDF lived lets you browsers unless you deal with Linked Data and in Python and RDF live web let's you create a really simple
browser to explorer so
this is running on my local machine are implemented it and about 4 lines of Python using RDF for the Web and it shows you all I can look through what a chunk is has these different properties and then I can look through all the chunks that we have defined in the reference definitions um his C Python and that defines and configure commands for example and then it shows me the linkage between them so that gets
referred to in a few different strata for example so my point is that this is really easy to do once build instructions a represented a simple young files or stored in the database you can reuse analysis tools like this which has not developed at all for build tools that is a general-purpose thing and we can now use it for analyzing build instructions and I'd like to generate and interesting graphs in future as well have been to a lot of data visualization talks yesterday in interested of making pretty graphs and Natalie diagrams now so the final part of
the talk is how this can be useful for Python development and how many people use virtualenv civility lens really useful quite simple way of isolating a Python dependencies as a few problems which is if you 1 install library that needs a system library and you have installed a new system is nothing virtualenv can do about that so you can use the base structure willing to build a container which tracks all of the dependencies the need rather than just the Python ones I would recommend if you have a problem of virtualenv keep using it because it's much more convenient but if you find yourself reaching the limits of what vigilant canoe and finding actually have start installing packages and tracking dependencies elsewhere basically gives you a way of defining everything all the poison dependencies policy library dependencies and right down to the toolchain used to build the and train definitions by hand is a bit boring so we have a tool called the import tool which can import metadata from other packaging systems
and we developed a way of importing information from pi pi circle I love what went into this quite a lot of research on my colleagues and we try it's looking at the service repos Python projects and using a patch vision of picked to analyze what's dependencies and expressed but it's actually quite difficult to get information that way the problem is again because set up the y is constant we're were going to do anything that and so you find that adopt the ways that don't make sense to pick when you in the Repo so what we what we've ended up doing is we have a solution which sensible of virtualenv and environment users but to install the package and then uses decrease together list of dependencies it's not the most efficient solution because you have to compile any embedded C extensions of the things but as the hundreds that always works the very moment that see this so yeah but the idea is to
generate something which can be using a tool it's useful outside of Python libraries so I can show you how to
name a package that can it working if I have an internet connection I can interest 1 Alex amount so I found that some packages some that you expect to have a lot of dependencies done actually list any for example Django and known by Douglas the dependencies in a machine readable way medicine and read me something that this so I guess is as good a Krylov compilation so actually especially um OK will
find a balance of bodies why we doing this is a few reasons
1 is that packing operating systems quite fun tracking operating system to link and what is that it does a lot of best practices today which some people follow and some don't and refine cells all of the time cleaning up after the after in projects where the best practices have been followed so making tooling where you can't actually avoids following best practices is a goal in some of these are not depending on third-party hosting most built systems today download tarballs from the upstream websites which is great and so website disappears or gets compromised recently tutorial store grown offline for example forever and all of the source code mirrored on notorious disappeared which would be really annoying except that we've been mirroring all of the projects we needed for years anyway and so it didn't make much difference at some point we have to find the new extremes for the ones that have made so they keep up to date but you can imagine if you if you have a build system which plans to from the torus and then the day before you really it disappears that's a real problem I Thiobacillus Maria it you know that you insulated from that and so that making missiles there's really is using the trove several clients trusting third-party binaries is another thing which is becoming really common at the moment with the rise of Dukkha which is great right downloaded binary which you can't released inspector source for religious roots in your computer and in a bunch of namespaces now so we build things from source instead you that's what we wanna write tooling which builds everything from so so you don't have to trust Running binaries downloaded from the Internet and 2 other things giving means up to date and making it as easy as possible to fix them up upstream so because everything's in gates you can follow any any component that you think is there is a problem and you can click on it straight away from local so that you have to worry about what format it's in anything else and then when she worries about the problem is you can analyze data transmitted to the project but it we discourage and patching things In the build instructions a lot distributions carried endless patches against projects which never seem to get up streams and some of them can't be some of them as legitimate things which destroyed specific but really wanted discourage patching things because it makes it more difficult than you come to upgrade from Poisson 3 . 4 apart and 3 . 5 and it turns out of your patch is no longer apply and so you don't upgrade for a long time to encourage building things directly from Cisco and that's so I want to talk about some things like from a study of the happy to take any questions FIL B follow thank you have a lot of patience in fills so going to compare system despite with what sort of
what you might get a particular and I yes I have used the other packets starts taking an image that's already built so I'll take a saner upon to based image and then it runs i convertible into different commands like you can run the shaft and mineshaft script or enhanceable enhanceable scripts and then they can deploy the image someone so it's in a related area and they overlap I actually thought of 1 point about writing the base struck plugin for packet which kids instead of starting with a to image which start by building or using a cached version of a system from code so the answer is they can they can interoperate at the moment they don't but I'd like to look at how to integrate patient with Parker since the down looks looks along and also on the UNIX systems and the tooling only works on what would be the words in any parse existence most of some of the tools and the work in basic itself to free us from and to track dependencies and make a distributions so Linux positive and the mother convenience of some market in the schools like a rocket or wider all exceed going to those no use what was system and not at the moment no bloody interested in implementing that some books in library if you want a lot of helpful links great yeah and so on last questions symbolic the what next baggage manager next less yes that's an excellent question I do know about Nick's OS and thing it's a great project I'm terrified of complexity by over a much larger line everything we do doing with them as as it becomes possible it thank you and I'm a bit slow between years so forgive me you probably already addresses do I understand correctly that with base rock like a do a sort of Gen 2 types of some the entire system is built from sources but there's no way then I can start from my consistent also are wbs that's correct yeah that's that's it yeah OK so that the actor integration would be pretty also probably I should write that myself thank you 88 I so very separately this spilled is happening is that the true figure censuring is commenced the highest and true new messages an integration thing that happened afterward if you're 1 of perform modifications of things in the true imagining that I'll post and post install commands yeah so the post and still commands basically those exist so that you don't have to override the default install commands so for example for autotools the default is make install this is not actually having to execute any commands inside of the truth itself yeah there's commands will run inside the jury but some them wondering how do you deal with architecture differences for example and things of that nature where you there your build host doesn't support the target like running into fuels executables inside the target so just compiling yes 1 example yeah that doesn't support cross compiling deliberately to avoid the complexity of supporting cross compiling I yeah there's a whole other set a series of ever run into a similar like as Linux is another example where were you will host dozens support across Major kernel versions that have and we recommend running builds inside a base rock the troops and so the only thing that effects of this kernel versions so there is a requirement on what can we have the can get around that by using a research on you same target and all those essentially yeah yeah called hi and set a couple questions how hard he reached the slide where do you get made from that's a good question the biggest upsets in quite interesting it's based on the Linux from scratch bootstrap and if you want to see the gory details and you can look in the definitions repository a which is and get server and the
gist of it is that we start by building from tarballs rebuild we have a bootstrap build mode which happens outside the jury and uses
the host tools so that builds um I think that builds the GCC and been utils and then with that it builds a stage 2 which is 6 components I think make GCC BusyBox and she would say and then it builds everything again with those tools in Ituri basic level ordering and the actual that the
description of this is in here so it's kind of
explaining comments and
she's starts at this stage when the new utils state and GCC and then Linux API had a students and so on In the
best that's quite good because it is really easy to cross bootstrap to a new platform so they search been ported to a bunch of different architectures ready like armed and we did it on the Indian port which I think this was the only OS is you can run on at the moment because it's you only need to cross build about 6 things and then the rest you can native build OK and so you have a follow-up question the if there is a security vulnerability and say G. steal or something like a level which he I should implication as you'd have to rebuild most of your image there is a way that you can cheat by adding the new company new version of the component on top so if you wanted you could add do you would see again an override the existing vision and deployed as an upgrade and we the design of it encourages rebuilding everything from service which is an ideal and and security updates you love compile machinery the more of us providing distributed build workers the better in the scenario
yeah so we use makes and great any clearly trying to fix the example of problems that they are trying to fix and using similar components and the way you aware of makes when you started your project these at I was yeah I was not the 1 of the founders of a shock I have kind of involved in a later on I was aware of mix of never really used it much found that has quite area of complexity laughter that build definitions rather than being data so functional codes and so I think long-term definitely different to align projects but didn't use next because he was scared of In a way yeah the people originally up by short and all think of using it the system has been developed in parallel OK I should add part of part of the original goal is basic is to reduce and by the Ministry of XML in the reality is generated starts in which I was always
allowances from site and conservative in here the so it's
quite simple example in the end wasn't the most efficient solution but it worked and so it just saves you writing definitions by hand things were metadata there already exists is also important for Ruby gems um and and something else now is that the least useful bootable Linux distribution over as an error depends how much you are using our XML from the consul
and the final version of the great while they were alleged to have
Distributionstheorie
Software
Physikalisches System
Lucas-Zahlenreihe
Distributionstheorie
Bit
Punkt
Prozess <Physik>
Dokumentenserver
Desintegration <Mathematik>
Browser
Versionsverwaltung
Dienst <Informatik>
Extrempunkt
Code
Raum-Zeit
Computeranimation
Task
Metropolitan area network
Physikalisches System
Bildschirmmaske
Prognoseverfahren
Code
Operations Research
Rippen <Informatik>
Gerade
Widerspruchsfreiheit
Gammafunktion
Hardware
Schnittstelle
Prototyping
Caching
Binärcode
Interpretierer
Dokumentenserver
Gebäude <Mathematik>
Softwarepiraterie
Physikalisches System
Quellcode
Gerade
Digitale Photographie
Portscanner
Arithmetisches Mittel
Quader
Dienst <Informatik>
Diskrete-Elemente-Methode
Menge
Flächeninhalt
Mereologie
Datenerfassung
Server
Dateiformat
Projektive Ebene
Größenordnung
Informationssystem
Standardabweichung
Chipkarte
Maschinenschreiben
Dualitätstheorie
Euler-Winkel
Dokumentenserver
Formale Sprache
Gruppenkeim
Dienst <Informatik>
Extrempunkt
Raum-Zeit
Computeranimation
Physikalisches System
Metropolitan area network
Code
Mustersprache
Speicherabzug
Visualisierung
Programmbibliothek
Radikal <Mathematik>
Skript <Programm>
Große Vereinheitlichung
Gammafunktion
Dokumentenserver
Gebäude <Mathematik>
Güte der Anpassung
Mailing-Liste
Quellcode
Physikalisches System
Speicherbereichsnetzwerk
Portscanner
Rechenschieber
Patch <Software>
Bildschirmmaske
Diagramm
Diskrete-Elemente-Methode
Menge
Mereologie
Projektive Ebene
Ordnung <Mathematik>
Chatbot
Instantiierung
Standardabweichung
Chipkarte
Dualitätstheorie
Versionsverwaltung
Dienst <Informatik>
Extrempunkt
Computeranimation
Systemprogrammierung
Spezialrechner
Metadaten
Metropolitan area network
Physikalisches System
Minimum
Zusammenhängender Graph
Spielkonsole
Bildgebendes Verfahren
Normalvektor
Gammafunktion
Quellcode
Physikalisches System
Elektronische Publikation
Speicherbereichsnetzwerk
Kreisbogen
Diskrete-Elemente-Methode
Menge
Server
Programmierumgebung
Repository <Informatik>
Verzeichnisdienst
Perpetuum mobile
Chipkarte
Morphismus
Bit
Prozess <Physik>
Momentenproblem
Nabel <Mathematik>
Desintegration <Mathematik>
Compiler
PASS <Programm>
Euler-Winkel
Raum-Zeit
Computeranimation
Übergang
Metropolitan area network
Code
Total <Mathematik>
Nichtunterscheidbarkeit
Skript <Programm>
Funktion <Mathematik>
Caching
Managementinformationssystem
Lineares Funktional
Namensraum
Datentyp
Hardware
Dokumentenserver
Datennetz
Gebäude <Mathematik>
Systemaufruf
Programmierumgebung
Quellcode
Ein-Ausgabe
Quader
Verknüpfungsglied
Beweistheorie
Grundsätze ordnungsmäßiger Datenverarbeitung
Server
Ordnung <Mathematik>
Verzeichnisdienst
Summand
Ext-Funktor
Sichtbarkeitsverfahren
Folge <Mathematik>
Subtraktion
Existenzaussage
Ablöseblase
Dienst <Informatik>
Code
Physikalisches System
Virtuelle Maschine
Mailing-Liste
Multiplikation
Front-End <Software>
Mini-Disc
Morphismus
Programmbibliothek
Zusammenhängender Graph
Datenstruktur
Bildgebendes Verfahren
Gammafunktion
Hardware
Binärcode
Softwarewerkzeug
Plug in
Physikalisches System
Modul
Chipkarte
Integral
Mapping <Computergraphik>
Flächeninhalt
Softwareschwachstelle
Meter
Caching
Parametersystem
Bitrate
Morphismus
Folge <Mathematik>
Subtraktion
Nabel <Mathematik>
Momentenproblem
Mathematisierung
Formale Sprache
Formale Grammatik
Gruppenkeim
Mathematische Logik
Code
Computeranimation
Eins
Physikalisches System
Metropolitan area network
Programmbibliothek
Skript <Programm>
Zusammenhängender Graph
Datenstruktur
Konfigurationsraum
Informatik
Stochastische Abhängigkeit
Hardware
Metropolitan area network
Caching
Umwandlungsenthalpie
Binärcode
Namensraum
Graph
Datennetz
Gebäude <Mathematik>
Datenmodell
Strömungsrichtung
Physikalisches System
Bitrate
Grundeinheit <Mathematik>
Quick-Sort
Digitale Photographie
Dienst <Informatik>
Menge
Rationale Zahl
Deklarative Programmiersprache
Mereologie
Wort <Informatik>
Speicherabzug
Instantiierung
Morphismus
Subtraktion
Elektronische Publikation
Dokumentenserver
Booten
Systemaufruf
Mailing-Liste
Physikalisches System
Computeranimation
Physikalisches System
Systemprogrammierung
Verzeichnisdienst
Softwaretest
Digitale Photographie
Mini-Disc
Speicherabzug
Programmbibliothek
Skript <Programm>
Klumpenstichprobe
Versionsverwaltung
Metropolitan area network
Physikalisches System
Punkt
Interaktives Fernsehen
Datenanalyse
Nichtunterscheidbarkeit
Browser
Softwarewerkzeug
Zusammenhängender Graph
Konfigurationsraum
Computeranimation
Morphismus
Kategorie <Mathematik>
Browser
Singularität <Mathematik>
Browser
Computeranimation
Spezialrechner
Virtuelle Maschine
Metropolitan area network
Physikalisches System
Systemprogrammierung
Benutzerbeteiligung
Interaktives Fernsehen
Standardabweichung
Wirtschaftsinformatik
Programmbibliothek
Konfigurationsraum
Gerade
Gammafunktion
Subtraktion
Punkt
Datenhaltung
Gebäude <Mathematik>
Ungerichteter Graph
Extrempunkt
Elektronische Publikation
Computeranimation
Metropolitan area network
Graph
Diagramm
Eigentliche Abbildung
Statistische Analyse
Visualisierung
Analysis
Managementinformationssystem
Kreisfläche
Wellenpaket
Momentenproblem
Mailing-Liste
Baumechanik
Physikalisches System
Digitale Photographie
Eins
Metropolitan area network
Metadaten
Patch <Software>
Dienst <Informatik>
Programmbibliothek
Inverser Limes
Projektive Ebene
Information
Maßerweiterung
Datenstruktur
Softwareentwickler
Repository <Informatik>
Maschinelles Sehen
Programmierumgebung
Einfach zusammenhängender Raum
Morphismus
Compiler
Mailing-Liste
Extrempunkt
Computeranimation
Internetworking
Digitale Photographie
Virtuelle Maschine
Metropolitan area network
Software
Programmbibliothek
Cloud Computing
Versionsverwaltung
Lucas-Zahlenreihe
Distributionstheorie
Subtraktion
Web Site
Punkt
Momentenproblem
Zellularer Automat
Patch <Software>
Binärcode
Computeranimation
Eins
Internetworking
Metropolitan area network
Streaming <Kommunikationstechnik>
Mailing-Liste
Weg <Topologie>
Client
Standardabweichung
Torus
Zusammenhängender Graph
Speicher <Informatik>
Binärdaten
Beobachtungsstudie
Namensraum
Rootkit
Gebäude <Mathematik>
Web Site
Physikalisches System
Quellcode
Binder <Informatik>
Digitale Photographie
Summengleichung
Patch <Software>
Verknüpfungsglied
Ablöseblase
Dateiformat
Projektive Ebene
Extreme programming
Distributionstheorie
Morphismus
Bit
Subtraktion
Punkt
Momentenproblem
Dokumentenserver
Natürliche Zahl
Bootstrap-Aggregation
Adressraum
Versionsverwaltung
Komplex <Algebra>
Code
Computeranimation
Kernel <Informatik>
Metropolitan area network
Dämpfung
Datenverwaltung
Syntaktische Analyse
Existenzsatz
Datentyp
Programmbibliothek
Skript <Programm>
Installation <Informatik>
Default
Figurierte Zahl
Bildgebendes Verfahren
Hilfesystem
Gerade
Soundverarbeitung
ATM
Gebäude <Mathematik>
Reihe
Indexberechnung
Plug in
Quellcode
Physikalisches System
Binder <Informatik>
Quick-Sort
Integral
Rechenschieber
Flächeninhalt
Projektive Ebene
Wort <Informatik>
Computerarchitektur
Message-Passing
Morphismus
Softwarewerkzeug
Indexberechnung
Dienst <Informatik>
Arithmetischer Ausdruck
Computeranimation
Übergang
Metropolitan area network
Deskriptive Statistik
Physikalisches System
Diskrete-Elemente-Methode
Endlicher Graph
ATM
Speicherabzug
Zusammenhängender Graph
Klumpenstichprobe
Versionsverwaltung
Subtraktion
Morphismus
Momentenproblem
Bootstrap-Aggregation
Versionsverwaltung
t-Test
Regulärer Ausdruck
Extrempunkt
Systemplattform
Computeranimation
Übergang
Metropolitan area network
Physikalisches System
Code
Zoom
Booten
Maschinelles Sehen
Bildgebendes Verfahren
Peer-to-Peer-Netz
Funktion <Mathematik>
Data Encryption Standard
REST <Informatik>
Computersicherheit
Machsches Prinzip
Softwarewerkzeug
Indexberechnung
Dienst <Informatik>
Softwareschwachstelle
ATM
Computerarchitektur
Compiler
Aggregatzustand
Metropolitan area network
Web Site
Erhaltungssatz
Flächeninhalt
Mereologie
Mixed Reality
Regulärer Ausdruck
Projektive Ebene
Zusammenhängender Graph
Physikalisches System
Komplex <Algebra>
Metropolitan area network
Distributionstheorie
Filetransferprotokoll
Morphismus
Booten
Systembereichsnetz
Versionsverwaltung
Reelle Zahl
Fehlermeldung

Metadaten

Formale Metadaten

Titel Introduction to Baserock
Serientitel EuroPython 2015
Teil 09
Anzahl der Teile 173
Autor Thursfield, Sam
Lizenz CC-Namensnennung - keine kommerzielle Nutzung - 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 10.5446/20199
Herausgeber EuroPython
Erscheinungsjahr 2015
Sprache Englisch
Produktionsort Bilbao, Euskadi, Spain

Technische Metadaten

Dauer 35:28

Inhaltliche Metadaten

Fachgebiet Informatik
Abstract Sam Thursfield - Introduction to Baserock The Baserock project is about creating system images from source code in a clean, reproducible way. All of the tooling is written in Python. In this talk I'll explain a bit about the core idea of Baserock: declarative system definitions (expressed in YAML) that can be built and deployed in various ways. Then I'll go into more detail about the tools available, and some of the cool things that they can do: distributed building, atomic system updates, creating custom container images, and more.
Schlagwörter EuroPython Conference
EP 2015
EuroPython 2015

Zugehöriges Material

Ähnliche Filme

Loading...