apache/mesos
Mesos
Apache Mesos is a distributed systems kernel and cluster resource manager that abstracts CPU, memory, and storage across a pool of nodes. It functions as a distributed infrastructure orchestrator, providing a layer to run multiple orchestration frameworks on a shared set of physical or virtual machines.
The system acts as a resource isolation engine, dividing a shared cluster into isolated containers to run diverse workloads concurrently. It enables multi-framework orchestration, allowing different distributed application frameworks to share a single infrastructure to maximize hardware utilization.
The project covers large scale compute distribution and distributed cluster management. Its capabilities include managing distributed resources and isolating compute power across multiple applications to prevent interference and ensure stable performance on shared servers.
Features
- Distributed Computing Frameworks - Provides a distributed infrastructure for running multiple computing frameworks across networked machines.
- Two-Level Scheduling - Implements a two-level scheduling mechanism that delegates task launching to individual frameworks.
- Offer-Based Allocation - Distributes cluster resources by offering available CPU and memory to frameworks for acceptance or rejection.
- Multi-Framework Orchestration - Supports running several different orchestration frameworks on a single shared infrastructure.
- Distributed Cluster Management - Manages a pool of shared computing resources across multiple nodes to ensure efficient hardware utilization.
- Distributed Infrastructure Orchestration - Provides a layer for running multiple orchestration frameworks on a shared set of physical or virtual machines.
- Workload Isolation - Divides a shared cluster into isolated containers to run diverse workloads concurrently without interference.
- Resource Isolation - Partitions system resources across applications to prevent interference and maintain stable performance.
- Heartbeat Monitors - Monitors worker node availability and cluster health via periodic heartbeat signals between master and agents.
- Distributed Systems Kernels - Abstracts CPU, memory, and storage across a pool of nodes to function as a distributed systems kernel.
- Namespace-Based Isolation - Implements resource isolation using Linux kernel namespaces and control groups to prevent application interference.
- Cluster Resource Managers - Acts as a cluster resource manager that provisions and scales compute power across diverse infrastructure.
- Cluster Resource Isolation - Provides strict isolation of compute resources to allow multiple frameworks to run concurrently on shared hardware.
- Multi-Framework Orchestration - Enables multiple orchestration frameworks to share a single physical infrastructure to maximize hardware utilization.
- Replicated State Machines - Maintains consistent cluster state across multiple master nodes using a distributed consensus algorithm.
- Framework Registrations - Allows different orchestration engines to connect to the cluster runtime dynamically without restarting the master.
- Distributed Compute Environments - Provides a distributed compute environment for executing workloads across a vast network of servers.
- Container Orchestration - Distributed systems kernel for resource management.
- Workload Orchestration and Scheduling - Distributed systems kernel for resource abstraction.
