|
|
Server for Continuous Integration
Šajdík, Michal ; Fiedor, Tomáš (oponent) ; Lengál, Ondřej (vedoucí práce)
This work contains description about the following topics: what kind of technologies and principles are needed for creation of a continuous integration server, already existing solutions, why there is a need to create a new one, and how to integrate continuous integration server which was created during this work, based on the information mentioned in this work, to a working environment. This work also shows effects and some side effects of correct and incorrect configuration of the mentioned continuous integration server. Mentioned continuous integration server is also able to run on MS Windows 10 and Linux without need to adapt a configuration for a specific operating system.
|
|
|
Kubernetes Canary Deployment Controller
Malina, Peter ; Burget, Radek (oponent) ; Rychlý, Marek (vedoucí práce)
The need to deliver value to users sooner rises each year in the competitive IT market. Agility and DevOps are becoming critical aspects of software engineering, seeking tools to enable and improve agile culture. Software projects in such culture tend to deal with new deployment strategies to reduce the risk of new changes breaking the existing system. However, staging and test environments almost always differ from the production. Using the appropriate deployment strategy, such as the canary, improves the overall stability of the system by first testing the new changes on a small subset of production traffic. Multiple experiments were made to prove that canaries can positively enhance deployment stability and reduce the risk that new changes bring.
|
|
|
Infrastructure for Testing and Deployment in the Field of Containers
Ormandy, Adam ; Lengál, Ondřej (oponent) ; Turoňová, Lenka (vedoucí práce)
Efficiency loss caused by repetitive manual tasks is a common problem throughout the IT sector. Developers often test, build, and deploy their software manually. That is not only time-consuming, but also dull and prone to errors and mistakes. This thesis tries to solve that in the context of one DevOps team, by unifying the development and testing tools, and by applying the principles of continuous integration and continuous deployment in the production environment. It is focused on Python, Jenkins, and container-based software and workflows. The main tools used in the thesis are GitLab CI, OpenShift and Tox. Thanks to work in described in the thesis, the number of projects with CI/CD pipelines increased from 7 to 50 percent, the amount of Python style violations started to decrease, containers have proper metadata, the container build process is automated, time and effort are saved by not doing repetitive tasks, and more.
|
|
|
Bulk Operation Orchestration in Multirepo CI/CD Environments
Víšek, Jakub ; Pluskal, Jan (oponent) ; Koutenský, Michal (vedoucí práce)
The multirepo model, where solution code is managed in many separate source control repositories instead of a single one, is gaining traction in software engineering. Amongst the disadvantages of this strategy is the amount of mundane work prone to human error involved in performing bulk operations, especially when these repositories are diverse in structure and utilized technologies. This thesis aims to design and implement a solution focused on time-saving and convenience of use that will allow for the definition and orchestration of development processes concerning many separate source control repositories. Finally, the completed solution is piloted in the production environment and evaluated.
|
|
|
Využití Nix/NixOps pro průběžnou integraci a nasazení software při vývoji
Vlk, Tomáš ; Burget, Radek (oponent) ; Rychlý, Marek (vedoucí práce)
Tato práce se zabývá uplatněním funkcionálního balíčkovacího systému Nix a jeho ekosystému (NixOS, NixOps) pro CI/CD při agilním vývoji. Při použití těchto technologií jsou problémy způsobené odlišným prostředím prakticky eliminovány bez nutnosti kontejnerizace. Práce obsahuje popis možností a nedostatků Nix/NixOps a navrhuje obecný postup použití těchto technologií pro jednotlivé fáze agilního vývoje a CI/CD. Díky Nix/NixOps je implementace CI/CD velmi jednoduchá a celý proces je navíc reprodukovatelný. Výstupem práce je sada příkladů demonstrující použití Nix/NixOps v různých projektech, která je dostupná jako open-source. Díky této sadě mohou vývojáři použít Nix rychle a jednoduše v jakémkoliv projektu, bez nutnosti studia velkého množství materiálů.
|
|
|
Bulk Operation Orchestration in Multirepo CI/CD Environments
Víšek, Jakub ; Pluskal, Jan (oponent) ; Koutenský, Michal (vedoucí práce)
The multirepo model, where solution code is managed in many separate source control repositories instead of a single one, is gaining traction in software engineering. Amongst the disadvantages of this strategy is the amount of mundane work prone to human error involved in performing bulk operations, especially when these repositories are diverse in structure and utilized technologies. This thesis aims to design and implement a solution focused on time-saving and convenience of use that will allow for the definition and orchestration of development processes concerning many separate source control repositories. Finally, the completed solution is piloted in the production environment and evaluated.
|
|
|
Využití Nix/NixOps pro průběžnou integraci a nasazení software při vývoji
Vlk, Tomáš ; Burget, Radek (oponent) ; Rychlý, Marek (vedoucí práce)
Tato práce se zabývá uplatněním funkcionálního balíčkovacího systému Nix a jeho ekosystému (NixOS, NixOps) pro CI/CD při agilním vývoji. Při použití těchto technologií jsou problémy způsobené odlišným prostředím prakticky eliminovány bez nutnosti kontejnerizace. Práce obsahuje popis možností a nedostatků Nix/NixOps a navrhuje obecný postup použití těchto technologií pro jednotlivé fáze agilního vývoje a CI/CD. Díky Nix/NixOps je implementace CI/CD velmi jednoduchá a celý proces je navíc reprodukovatelný. Výstupem práce je sada příkladů demonstrující použití Nix/NixOps v různých projektech, která je dostupná jako open-source. Díky této sadě mohou vývojáři použít Nix rychle a jednoduše v jakémkoliv projektu, bez nutnosti studia velkého množství materiálů.
|
|
|
Server for Continuous Integration
Šajdík, Michal ; Fiedor, Tomáš (oponent) ; Lengál, Ondřej (vedoucí práce)
This work contains description about the following topics: what kind of technologies and principles are needed for creation of a continuous integration server, already existing solutions, why there is a need to create a new one, and how to integrate continuous integration server which was created during this work, based on the information mentioned in this work, to a working environment. This work also shows effects and some side effects of correct and incorrect configuration of the mentioned continuous integration server. Mentioned continuous integration server is also able to run on MS Windows 10 and Linux without need to adapt a configuration for a specific operating system.
|
|
|
Infrastructure for Testing and Deployment in the Field of Containers
Ormandy, Adam ; Lengál, Ondřej (oponent) ; Turoňová, Lenka (vedoucí práce)
Efficiency loss caused by repetitive manual tasks is a common problem throughout the IT sector. Developers often test, build, and deploy their software manually. That is not only time-consuming, but also dull and prone to errors and mistakes. This thesis tries to solve that in the context of one DevOps team, by unifying the development and testing tools, and by applying the principles of continuous integration and continuous deployment in the production environment. It is focused on Python, Jenkins, and container-based software and workflows. The main tools used in the thesis are GitLab CI, OpenShift and Tox. Thanks to work in described in the thesis, the number of projects with CI/CD pipelines increased from 7 to 50 percent, the amount of Python style violations started to decrease, containers have proper metadata, the container build process is automated, time and effort are saved by not doing repetitive tasks, and more.
|
|
|
Kubernetes Canary Deployment Controller
Malina, Peter ; Burget, Radek (oponent) ; Rychlý, Marek (vedoucí práce)
The need to deliver value to users sooner rises each year in the competitive IT market. Agility and DevOps are becoming critical aspects of software engineering, seeking tools to enable and improve agile culture. Software projects in such culture tend to deal with new deployment strategies to reduce the risk of new changes breaking the existing system. However, staging and test environments almost always differ from the production. Using the appropriate deployment strategy, such as the canary, improves the overall stability of the system by first testing the new changes on a small subset of production traffic. Multiple experiments were made to prove that canaries can positively enhance deployment stability and reduce the risk that new changes bring.
|
host ::
RSS.