This article was accepted into the corpus but its outbound wikilinks were never NER-processed — typical at the deepest BFS hop or when the run's entity cap was reached. No expansion funnel to show.
| Workload Manager (WLM) | |
|---|---|
| Name | Workload Manager (WLM) |
| Developed by | IBM, Microsoft, Amazon, Google, Red Hat |
| Initial release | 1980s |
| Latest release | ongoing |
| Operating system | z/OS, Linux, Windows, macOS |
| License | proprietary and open source |
Workload Manager (WLM) is a system-level service that controls, prioritizes, and schedules computational work across resources in mainframe, server, and cloud environments. It enforces policies to meet service-level objectives for applications from transactional systems to batch processing. WLM implementations coordinate with resource managers, hypervisors, orchestration tools, and monitoring platforms to deliver predictable performance.
Workload coordination traces to systems such as IBM z/OS, UNIX System V, Microsoft Windows NT, Linux Foundation, VMware ESXi, and Amazon Web Services where administrators define objectives, priorities, and constraints. WLM mediates among resources like CPU, memory, I/O, and networking while interacting with platforms including Kubernetes, OpenStack, Red Hat Enterprise Linux, Azure, and Google Cloud Platform. By mapping business goals from owners such as CIO and SRE teams to runtime controls, WLM aligns enterprise services with objectives established by stakeholders like CTO and SLA signatories.
Origins appear in mainframe research from institutions like IBM Research, with operational roots in products such as OS/360 and later z/OS Workload Manager extensions developed during interactions with clients including Deutsche Bank, JP Morgan Chase, and CERN. Parallel evolution occurred in academic projects at Massachusetts Institute of Technology, Stanford University, and Carnegie Mellon University which influenced schedulers in BSD and System V. Commercial influences include Microsoft Research work on Windows Server scheduling, Sun Microsystems contributions to Solaris, and virtualization advances by VMware, Inc. and Xen Project. Cloud-era designs were shaped by practices at Amazon.com, Google LLC, and Netflix, Inc..
Typical architecture comprises policy engines, schedulers, counters, collectors, and enforcement agents. Core components resemble designs from IBM zEnterprise models and borrow patterns from Apache Mesos and Kubernetes Scheduler. Policy stores use formats inspired by JSON, YAML, and management protocols related to Simple Network Management Protocol linked to telemetry solutions like Prometheus and Elastic Stack. Enforcement integrates with hypervisors developed by Intel Corporation and Advanced Micro Devices as well as container runtimes by Docker, Inc. and containerd.
Scheduling algorithms adapt concepts from Earliest Deadline First, Round-robin scheduling research, and priority inversion remedies from Priority Inheritance Protocol studies. Policy management borrows from standards and frameworks such as ITIL, COBIT, and enterprise governance from ISO/IEC 20000 and ISO/IEC 27001 where applicable. Administrators express service classes, weighting, capping, and affinity rules influenced by case studies from Goldman Sachs, Walt Disney Company, and Pfizer that require predictable throughput and latency.
WLM relies on telemetry collected via agents aligned with protocols from OpenTelemetry and time-series databases such as Prometheus and InfluxDB. Key metrics parallel those studied in performance engineering at Bell Labs and reported in literature from ACM and IEEE. Dashboards often use visualizations from Grafana or Kibana and alerting integrates with platforms like PagerDuty and Slack (software). Performance indicators include throughput, latency, utilization, fairness, and anomaly rates referenced in research by DARPA-funded projects and academic conferences like USENIX and SIGMETRICS.
Enterprises apply WLM to transactional processing at institutions such as Bank of America and Visa Inc., scientific computing at Los Alamos National Laboratory and CERN, and media streaming at Netflix, Inc. and Spotify Technology S.A.. Implementations include IBM Workload Manager for z/OS, scheduler integrations in Kubernetes via Kube-scheduler plugins, resource controllers by Red Hat in OpenShift, and cloud-native offerings from AWS Auto Scaling and Google Kubernetes Engine. Hybrid scenarios combine on-premises stacks from Dell Technologies and Hewlett Packard Enterprise with cloud services from Oracle Corporation.
Interoperability depends on standards and adapters connecting to identity providers like Okta, Microsoft Azure Active Directory, and LDAP directories used at organizations such as NASA and Department of Defense (United States). Integration layers use APIs and protocols including gRPC, RESTful API, OAuth 2.0, and SAML to coordinate with CI/CD pipelines built with Jenkins, GitLab, and GitHub. Cross-vendor orchestration leverages patterns from Cloud Native Computing Foundation projects and consortiums involving OASIS and vendor alliances like OpenStack Foundation.
Category:Computer system software