Some Readme updates
This commit is contained in:
22
README.md
22
README.md
@@ -9,6 +9,7 @@
|
|||||||
|
|
||||||
## About Warewulf
|
## About Warewulf
|
||||||
|
|
||||||
|
### History
|
||||||
For over two decades, Warewulf has powered HPC systems around the world. From simple “under the desk” clusters to large
|
For over two decades, Warewulf has powered HPC systems around the world. From simple “under the desk” clusters to large
|
||||||
institutional systems at HPC centers as well as enterprises who rely on performance critical computing.
|
institutional systems at HPC centers as well as enterprises who rely on performance critical computing.
|
||||||
|
|
||||||
@@ -23,7 +24,22 @@ solution. But over the last few years, there have been many innovations in Enter
|
|||||||
should) be leveraged as part of Warewulf. Additionally, some of the lessons learned from Warewulf v3 architecture
|
should) be leveraged as part of Warewulf. Additionally, some of the lessons learned from Warewulf v3 architecture
|
||||||
should be rolled into an updated architecture for provisioning management.
|
should be rolled into an updated architecture for provisioning management.
|
||||||
|
|
||||||
At the core, Warewulf focuses on what has made Warewulf so widely utilized: simplicity, ultra scalable, lightweight,
|
### Warewulf v4
|
||||||
and easy to manage solution built for entry level system administrators to be able to design a highly functional and
|
Leveraging this legacy of provisioning and cluster management brings us to where we are today. The next generation
|
||||||
easy to maintain cluster no matter how big or small it is.
|
of Warewulf. Warewulf v4 is a complete rewrite in GoLang, taking in the legacy of what we've come to expect with
|
||||||
|
Warewulf, bringing it into the present, and looking out into the future.
|
||||||
|
|
||||||
|
Warewulf v4 combines ultra scalability, flexibility, and simplicity with being light weight, non-intrusive, and a
|
||||||
|
great tool for scientists and seasoned system administrators alike. Warewulf empowers you to scalably and easily manage
|
||||||
|
thousands of compute resources.
|
||||||
|
|
||||||
|
### Architecture
|
||||||
|
One of the design tenants of Warewulf is how to scalably administrate many thousands of compute nodes. Generally
|
||||||
|
speaking, operating system state introduce a surface for potential discrepancies and version creep between nodes and
|
||||||
|
thus Warewulf has always gone with the "single system image" approach to clustered operating system management.
|
||||||
|
This means that you can have a single management node "image". For the last 20 years, we've been calling this a
|
||||||
|
Virtual Node File System (VNFS), but in hindsight, this is a container.
|
||||||
|
|
||||||
|
At its core, Warewulf v4 focuses on what has made Warewulf so widely loved: simplicity, ultra scalable, lightweight,
|
||||||
|
and an easy to manage solution built for both scientists and seasons system administrators to be able to design a
|
||||||
|
highly functional yet easy to maintain cluster no matter how big or small or customized it needs to be.
|
||||||
|
|||||||
Reference in New Issue
Block a user