Logo TIB AV-Portal Logo TIB AV-Portal

Grocker, a Python build chain for Docker

Video in TIB AV-Portal: Grocker, a Python build chain for Docker

Formal Metadata

Grocker, a Python build chain for Docker
Title of Series
Part Number
Number of Parts
CC Attribution - NonCommercial - ShareAlike 3.0 Unported:
You are free to use, adapt and copy, distribute and transmit the work or content in adapted or unchanged form for any legal and non-commercial purpose as long as the work is attributed to the author in the manner specified by the author or licensor and the work or content is shared also in adapted form only under the conditions of this license.
Release Date

Content Metadata

Subject Area
Fabien Bochu - Grocker, a Python build chain for Docker Grocker is a Docker build chain for Python. It transforms your Python package into a self-contained Docker image which can be easily deployed in a Docker infrastructure. Grocker also adds a Docker entry point to easily start your application. ----- At Polyconseil, we build Paris electric car sharing service: Autolib'. This system is based on many services developed using web technologies, Django and our own libraries to handle business logic. Packaging is already a difficult problem, deploying large Python projects is even more difficult. When deploying on a live and user- centric system like Autolib', you cannot rely on Pip and external PyPI servers which might become unavailable and are beyond your control. In the beginning we used classic Debian packaging: it was a maintenance hell. It took hours to build our packages and update their metadata to match our Python packages. So we switched to Docker. Docker allows us to have a unique item that is deployed in production systems: code updates are now atomic and deterministic! But before deploying the Docker image, you need to build it. That's where Grocker comes in. Grocker is a Docker build chain for Python. It will transform your Python package into a self-contained Docker image which can be easily deployed in a Docker Infrastructure. Grocker also adds a Docker entry point to easily start your application.
Context awareness Installation art Run time (program lifecycle phase) Distribution (mathematics) INTEGRAL Multiplication sign Source code Design by contract Mereology Food energy Computer programming Web 2.0 Medical imaging Mathematics Web service Different (Kate Ryan album) Extension (kinesiology) Physical system Adventure game Covering space Scripting language Predictability Instance (computer science) Flow separation Connected space Data management Message passing Process (computing) Unabhängigkeitssystem Order (biology) Resultant Spacetime Server (computing) Existence Wage labour Information systems Power (physics) Revision control Propagator Peripheral Bridging (networking) Natural number System programming Data structure Form (programming) Condition number Standard deviation Dependent and independent variables Focus (optics) Matching (graph theory) Server (computing) Cellular automaton Chemical equation Gender Variance Cartesian coordinate system Database normalization Computer animation Logic Information retrieval Musical ensemble Service-oriented architecture Pressure Library (computing)
Octahedron Length INTEGRAL Direction (geometry) Computer-generated imagery Interactive television Electronic program guide Set (mathematics) Compiler Medical imaging Object (grammar) Rootkit Information Extension (kinesiology) Condition number Physical system Data type Default (computer science) Predictability Area Distribution (mathematics) Online help Building Point (geometry) Basis <Mathematik> Database Bit Cartesian coordinate system Limit (category theory) Word Computer animation Formal grammar Website Software testing Summierbarkeit Physical system Reading (process) Local ring Library (computing) Extension (kinesiology)
Octahedron Pattern recognition Run time (program lifecycle phase) Online help Run time (program lifecycle phase) Building Multiplication sign Software developer Point (geometry) Interactive television Inference Medical imaging Computer animation Different (Kate Ryan album) Object (grammar) Convex set Configuration space Information Physical system Default (computer science) Data type
Source code Frequency Computer animation Run time (program lifecycle phase) Source code Configuration space Receiver operating characteristic Default (computer science)
Area Matching (graph theory) Computer file Multiplication sign Set (mathematics) Incidence algebra Revision control Medical imaging Goodness of fit Process (computing) Computer animation Bridging (networking) Different (Kate Ryan album) Operator (mathematics) Configuration space Key (cryptography) Musical ensemble Extension (kinesiology)
from really looking fault yeah soon pleasure it's book in which will shoot was the talk grow called book of built change for Python applications so he and as we will talk about broker is which and from the the the q he match from it and getting the kitchen some of the differences get some context are more considered but could say it is the company behind the retrieval performance of systems and that it is in the the contract so based in part in fact we have 5 cultural resources in the world but it is the 1st and the largest 1 so each other about the 1st of these properties and that it would be the 2 1 is a information system is composed of set education and want to education and do not include mechanism that and that is reduced set the concept the and so this condition honestly based on gender replication of the initial focus so there is and you're whether laborers to work on businesses and the problem was to the process of education in which and I have worked on develops subjects from time to time and why it into into all way we've been involved in different using go to the prior application prediction we use the indicator and didn't think that was found in tutors and 15 for the time applications you actually did you of that in the kitchen that at that time to is an there and through the repulsive versions of already declared in you it and it in your skin is really to filter had powers to the kids propagation on all these differences the founders of we we can refer you repetitions and so which is it's interesting that it's too is that sort if you take 2 days to a it is that it is something we have when we scrutinized foreign made its replication was about 6 times and I have reverted to the application is submitted to and you're edition processes going on in the habits of they that because we have physical devices like our judgment and logic careers and say that I have to be mini-database humans so we have to look the the girl has to have an etymon and I want you want to have the existence of an implication of we is that the support and manage and you can't have an uncertain standard at the dishes as energy the learners and as I mentioned so you if that will process space Europe prediction his family and good you have to have to would move education understand that cell of the from the only 1 application that is and most of the variance of evidence the idea is that our results will decide to do the same thing as go up that's only in a different way for isn't their adventures that system images do approximately the sensing that stuff from source and the between the gage and like what we started writing region knows that assists images rogard comes from many of our approach tried the 1st 1 was to use that and so on the predictions of that's an infected with and no we use is the pressure because this two-dimensional and resistance was who have to the extension in place and the predictions of balance in the face so the next approach was too improved rural rural didn't occasion to tourism and like this is from the was and so we get and the switch to declare approach this approach or you you have you and additional libraries increase your role addresses the programmer library so if you want to read that's 1 of the few libraries you have to be sure that you at edition bands and efforts of the program what using a intervention and and is very difficult you have to do is to changes in the sense that and in it the next especially tried was to use the cover that twisted-pair fairly have to work at breeder and several independent systems and repetition and remove from dependencies in 1 later on Monday I was a always the image his eyes will be a very big no because accepted because again connecting data values from the so but don't that it fails it just enough of them as they did it just much and so the rest of Gaussian we tried was a slogan which in fact is that they would like pushed in the variable is a very popular pass that summarizes some of those will not know In this is the approach we use a bridge there and the inferred as the speed is Britain's responses there is an edge which include value around runtime dependencies and redundancies you and since it's 1 instance of this image and you really during World applications you said is the result of the preprocessed and you did it is the stuff and then I want you run your application we lost steps in new instances in match what the result of the preprocessing and on somewhere around in Europe it in your instance that's the person of the program you have 1 big business natural order reputation so you cannot have around and education especially requirements all so you have a very big and fat image integral part of the things that cure us that structure with we use a surprise on 100 and so on so that the talks once you wouldn't entries order we stand by putting in Bayesian edge on the lookahead which is funded in distance and once it is really that adds you're runtime dependence is to create which he managed this image we been that we really use it much as best of the combined image which is that there is a major reason extended value the time dependence is gonna compare script this image be around to conveys a win for eradication and this and this is this and this is the really starting at the and then we reuse and what's the other 2 were exposed 1 of the web services web server is is when we will use the grids there were enough connection with his would you talk to the coverage forms or which imagine and incentives that will we have been precompiled and comparing it infects is not a context that this because we have over 1 were to imagine that the dependency of not wanting to 66 that's done matter is the best time to talk for a worker walks the
comparing image is to erode to compared with an explicitly 6 extension and then to the library and read that the instead of a new world which is the new him at on the other words detailed and they are problem gabbroic extraction we use it area day to deploy your application and prediction but they have current currently have some limitations this is 1 of them is that there there is an edge if there was tremendous so is that you can undergo guide formal justification is a very Europe this additional is that there are and is library you'll we chose to decode the tracker it was also the that and on education in the business of 6 and when it's a very big basis images on which image and maybe it's because he was that they which based image is 100 and 25 megabytes compared to a length which is has a distribution like techniques which enables users on the site given that's what they need to do not have said that army user-defined dying you prediction application another problem was the current condition is that it can only be in the kitchen education and in fact the direct sum F to be and honored by basically become the ability of a public database of them that has to be commentators from those of us who just on we currently working on it with the other 2 of them and then there is a In an effort to yes so I want to use it a
very simple you just have to work set of local bit unsettled and send me a white on your command line integral Grid medication name chuig assignment and reputation system and that there no sometimes will have a look at image of the reputation of users can Mr. Chairman
insight is any time often things that the recognition have to be taken and enhanced user there are at least be competitive with inference and we do not have ground I dependencies so it's a and is a good example although I we have a convex I said that we you have to accept with dependency and runtime dependence and religion answers to 1 of the things that are you can use to explain issue with
Jefferies is I use 1 of them I said there and tend to hydrogen so that you can see it is like a times is greater once you stands at the image by different and developed use an entrepreneur Corcoran that I didn't unhappy this so I grow grow
wasn't interested yesterday so you can see that is source on the frequency the period that it didn't recommend that engages in Baghdad and so that can be into so if you want to contribute to our Cervantes and the
artisans and the question is OK what is the size of your image but to a big issue a generative why is the size of the image such a big issue I mean am also although images but the way bigger than 6 million because we want to use the best path so if we if we do not know where's the image would be the put in your issue and of the few places that for some reason you have to pull of hundred you image if user but I want to get is a very long process if you have some ontology matching is a responding really fast the hi thanks for the story as you say was the wall dangerous and can you give us some more details you know it was a approach artists did uh 1 you would say extension you wouldn't know that issue as good but time dependences and set B we shouldn't consider the a very meticulous of halfway through and sentences from the economies of data or something like that and you should do is from the operations center z is a bridge process then the replication of not having a bearing on so it is not a matter more questions but that the config file you talk about what kind of formant that doesn't support the configuration file you you can supply areas they know what kind of performance those of is this is identified maybe I have to give an example for more which which it is a version of the underworld and between into uh to can verify fraudulent have edition is the so you can community began the sets you'll never really Europe Ivorians neurons front you your process and stand on the issue of Wednesday kinds of
documentation we on the on the bench so you can say you can put dependence is just something from differences can cities that you just saw then is because for example the user version that you should have the following kind of people that you do not have a band precedent extension is used users society or something like that and you can see other dependency where we're at time 1 and intervals already stuff that from that prevented him more questions we have time of look and incident thank