Sardina FishOS

Highly efficient Clouds for businesses with OpenStack


FishOS

Sardina FishOS is a revolutionary, highly reliable enterprise cloud platform that provides a turnkey infrastructure for traditional workload and next-gen applications. It provides a complete lifecycle view of the OpenStack cloud and Kubernetes Orchestration to enable enterprises to deploy securely, predictably, reliably, and without downtime. OpenStack delivers the best in cloud technology, serving as the framework for creating highly scalable clouds for the most demanding workloads. In combination with FishOS, this is the most efficient cloud platform available.


FishOS HCI - Hyperconvergent cloud platform

Mit With FishOS HCI , organizations can quickly and easily deploy a highly reliable cloud platform for traditional enterprise workload and next-gen applications. These include data security, privacy, zero-click operations, and easy-to-use AI capabilities through automation.

FishOS'award-winning AI-based decision engines deliver unmatched 10x TCO, elastic pay-as-you-grow economics, elimination of oversupply and predictive risk, and a smart, extremely high cost of ownership Reliability of 24x7x365.

Benifits of FishOS:

Rich in featuresFishOS comes standard with useful features that businesses need, including automatic failover, disaster recovery, smooth upgrade paths, monitoring, integration, and coexistence with existing infrastructure.
Highly scalableFishOS grows with your cloud - it's designed to scale over 100,000 physical servers.
Maximizes the utilizationWith unique technologies for tracking and optimizing data center resource utilization, FishOS is doubling and even trebling your server utilization rates.
Reduced TCOThe ability to gain insight into how your data center works, to ensure automated operations, and to quickly deploy and redistribute workloads allows FishOS to manage server states to reduce their total cost of ownership by a factor of 10.
Increases the ROIWith FishOS, you can dramatically improve the utilization of your server and get the most out of your hardware investment.
Optimizes the CapExBecause fewer servers are required to process workloads, the cost of hardware and equipment is reduced.
Smooth integrationFishOS integrates well with your existing infrastructure alongside your traditional environment. No replacement and no long integration time are required.

FishOS in HPC

OpenStack's private cloud enables HPC facilities to respond quickly and flexibly to different user needs. However, the stringent limitations and requirements of HPC present huge challenges in running a private cloud.

FishOS overcomes these challenges, resulting in lower operating costs and greater efficiency.

Challenges:

  • Traditional HPC environments limit users' operating environment and provide a high acceptance threshold for new users
  • Hardware failures affect long-running jobs and force users to restart calculations or implement a cumbersome and error-prone checkpointing process
  • Long-running jobs must be interrupted to allow urgent jobs to run or a complex, error-prone process to prepay the job
  • Because resources are statically managed based on reservation, HPC equipment operators have unused resources and users have a longer delay in calculation results
  • Operators of HPC systems can only respond to (reservation) abuse through long consultation procedures
Solutions:

  • With FishOS, HPC equipment can meet a wide range of user needs without affecting other users - users have their own private environment
  • FishOS provides dynamic error handling that preserves states of long-running calculation jobs without the need for lengthy and error-prone checkpointing processes
  • FishOS provides a flexible and reliable work instruction through dynamic management of compute resources
  • FishOS enables higher throughput by optimizing dynamic resource requirements to the resources available
  • FishOS eliminates abuse by managing resources based on live usage

Differences to standard OpenStack

It's one thing to get OpenStack running, a completely different one, to be scalable, reliable, efficient, and cost-effective. The key values of FishOS lie in these critical aspects of cloud operations.

Managing resources in a data center is essentially about optimally solving the problem of live demand relative to resource availability, constantly monitoring resource conditions, and rapidly implementing preventative and corrective actions. FishOS controls resources and makes decisions based on a wealth of detailed resource usage data across the entire cloud infrastructure to ensure that your OpenStack data center is highly functional and highly available at optimal cost.



  • Highest availability: Each component of FishOS has been designed to operate in high-availability mode, providing redundancy and fully guaranteeing the uptime of the service.
  • The integrity diagnostics framework tracks the state of the host server and can automatically migrate the workload to another host.
  • Best service: FishOS can scale across various types of operational automation that would otherwise require Admins intervention. This design allows a single operator to manage a data center with thousands of host servers.
  • Places VMs fast and optimally in a large, dynamically changing data center. VMs are immediately placed on the right host - for the first time and at any time. Waiting for the start of a VM falls.
  • Automatic redistribution of workload of VMs across the plant while meeting workload requirements to ensure optimal VM performance and maximize data center capacity at a constant TCO.
  • Number of servers will be adjusted to the size, to meet the workload requirements, which significantly reduces operational costs and optimizes CapEx for servers and assets.
  • Knowing the data center structure allows for the physical location of a host server to be considered when placing and balancing the workload.
  • Smooth upgrade: With FishOS, you do not have to worry about upgrading OpenStack. A secure and smooth upgrade path is guaranteed.

Container Virtualization

In addition to the classic virtual machines, the so-called Docker containers and the container orchestration can also be operated via Kubernetes on OpenStack. Both technologies have meanwhile revolutionized modern IT and are increasingly being noticed by operators and developers.

Docker - Isolation by application container

With the free software Docker applications can be packaged in containers, which in turn represent processes and are isolated from the rest of the system. The application will be composed of small blocks and represents the modern architectural approaches such as: immutable infrastructures, microservices and distributed software. The greatest value of Docker lies however in the image format and in the mechanism to distribute these. Regardless of the operating system, applications can be seamlessly deployed in different environments.

Docker

Kubernetes - Orchestration and automation of container applications

Although Kubernetes is not considered a traditional all-inclusive PaaS because it does not operate directly at the hardware level, it is a simple and extensible open source management platform of container workloads and dinstems, an optimal orchestration solution. The software provides a container-centric management environment that is decoupled from the infrastructure and the host file system and allows for easier and more efficient creation of container images, e.g. compared to the VM images. With Kubernetes, operators are able to simplify and automate the deployment, operation, maintenance and scaling of container-based applications. It groups groups of hosts running the containers into clusters of physical or virtual machines and manages them as a single entity.

Kubernetes