TULPAR is an enterprise control panel designed to manage HPC (high-performance computing) clusters through a web browser. It combines Slurm job management, Warewulf 4 provisioning, node health monitoring, infrastructure services (NTP/LDAP/DNS/MySQL) and Docker management in a single interface.
The backend is Python/FastAPI, the frontend React 18 + TypeScript. Metrics come from Prometheus and job information from the slurmrestd REST API; data lives in MariaDB and PostgreSQL/TimescaleDB. The whole stack is deployed with Docker Compose — head Slurm components run natively on the host.
No. Installation is two commands and a web wizard; node onboarding, service provisioning and configuration convergence are automatic. Researchers can submit jobs from a web form without writing sbatch scripts. Familiarity with Slurm concepts helps in advanced scenarios, of course.
Every license comes with comprehensive documentation — installation, administrator and user handbooks are embedded in the panel. With the Starter, Standard, Pro and Enterprise support plans we're with you from installation consulting to 24/7 operations support; use our contact page for details.
Yes. TULPAR is designed API-first with ~300 REST endpoints and works with industry-standard components such as Slurm, Prometheus, OpenLDAP, dnsmasq, chrony and MariaDB. Hybrid identity (local + LDAP/SSSD) with your existing LDAP directory is supported.
TULPAR is offered under a commercial license; licensing is annual, determined by your cluster scale and support coverage. Every license includes the full feature set — there is no feature gating. Contact us through the contact page for a quote.
Because metric collection uses the Prometheus API rather than SSH, it scales naturally to 100+ nodes. The per-node agents (node_exporter + tulpar-agent) are installed automatically during kickstart.
Yes. RPM repositories travel with the product, the Slurm package repository can be imported as a GPG-signed offline bundle, and container images are loaded with docker load. Everything works on a closed network, including installation, monitoring and central logging.
Node CPU/RAM/disk/GPU metrics, job accounting data, a unified event timeline, rule-based alerts (SMTP/webhook/in-panel notifications), a central log store (VictoriaLogs), and a job detail view joining Slurm accounting with Prometheus series.
TULPAR runs entirely on your own servers (on-premises) — your data never leaves your infrastructure. Every mutation is written to the audit log with SHA-256 hash-chain tamper detection; JWT-based authentication, RBAC and rate limiting are standard.
A typical installation takes minutes: the preflight script prepares the prerequisites, install.sh brings up the stack, and you enter your values in the web wizard. When the wizard finishes you're redirected to the login screen and the cluster is ready to use.