Please log in to watch this conference skillscast.
Both developers and operators view docker favorably, because it is suppose to make things easier. However, there is still effort required to manage the virtual infrastructure and machines in a distributed environment. Hyper_ provides a new container environment, where the users can launch containers on an “infinite” daemon.
There are several challenges in building a distributed docker daemon for multi-tenant usage, e.g. container isolation and network isolation between tenants, which the Hyper_ container technology combined with SDN has solved. Another challenge is storage, with a distributed docker daemon, the user's container may stop and start anytime, and may be scheduled to different physical hosts. Therefore, you must have a distributed storage engine for containers and images. In Hyper_, they built a multi-tenant and distributed graph engine for images and container rootfs based on Ceph, through which they can launch a container anywhere in the cloud.
The Call for Papers is now open for ContainerSched 2017! Submit your talk for the chance to join a stellar line-up of experts on stage. Find out more.
YOU MAY ALSO LIKE:
- FullStack Bytes with Jolyon Russ! (in London on 25th October 2017)
- F# eXchange 2018 (in London on 5th - 6th April 2018)
Hyper_: Build a Distributed Docker Daemon
Xu Wang is CTO and Cofounder of Hyper_. Prior to Hyper_, Xu served as the chief architect for China Mobile BigCloud, where he lead the kernel, distributed storage, and big data product lines. Xu has over 10 years of experiences in Debian, Linux kernel, Virtualization, Distributed Filesystem and NoSQL databases. He holds a PhD degree from Beijing University of Posts and Telecom (BUPT).