Doker (Doker)

Doker (Doker) #

One of many projects that support container technology
Container technology existed before, but it became widely known because of Doker.
De facto standard for container technology
Ranked 2nd most popular cloud open source in 2014, announced by the Linux Foundation
Usable on various operating systems
Simplifies building, deploying, and running by packaging not only applications but also dependencies and file systems.
Virtualization using kernel features like Linux namespaces and cgroups

Doker can be used with various cloud service models.

Image: A single file created after installing necessary programs, libraries, and source code.
Container: A virtual environment created by isolating and running an an image in an independent space.

Containers Solve It! #

Software components running on the same system often conflict or have various dependencies.
Containers are a technology that isolates each microservice using virtual machines.
Containers can run very fast because they don't implement all hardware like virtual machines.
If a process issue occurs, the entire container must be adjusted, so it's best to run a single process per container.

By utilizing space not needed by the hypervisor, more resources can be invested in applications.

Container Performance Comparison #

GFLOPS (GPU Floating point Operations Per Second) for Native, VM, and Container

GFLOPS (GPU Floating point Operations Per Second) is a unit primarily used to numerically represent computer performance.

Technologies for Container Isolation #

Linux Namespaces: Provides each process with an independent view of the system for file system mounts, networks, users, hostnames, etc.

Linux Control Groups: Limits the amount of resources a process can consume.

Limitations of Doker #

As services grow, the number of containers to manage increases rapidly. Even with Doker, management becomes difficult, and deployment and container placement strategies like scale-in and scale-out are challenging.