Slurm Services — Installation, Consulting & Managed Support
Enterprise Slurm workload manager installation, configuration, consulting, and 24/7 managed support services. Turkey's HPC solutions partner Mevasis.
Slurm is the open-source workload manager used by the majority of HPC clusters worldwide. Trusted by NASA, CERN, and TOP500 supercomputers, Slurm excels when properly configured — but a misconfigured Slurm installation leads to resource waste, unfair queue distribution, and user dissatisfaction.
Mevasis has been delivering enterprise Slurm installation, optimization, and managed operations for 12 years across Turkey. We don’t just install the software — we design scheduling policies tailored to your workloads and user profiles.
Slurm Installation Services
For new HPC clusters or integrating Slurm into existing infrastructure:
- MUNGE authentication — encrypted inter-node communication with OpenLDAP integration
- Partition architecture — short/long/GPU/debug queues matched to your workload mix
- Cgroup resource isolation — CPU, memory, and swap limits to protect co-located jobs
- GPU GRES configuration — NVIDIA A100/H100 GPU scheduling with proper device mapping
- Fair-share and quota policies — department/project-based resource distribution
- License management — ANSYS, LS-DYNA, MATLAB license integration
- Prometheus + Grafana monitoring — real-time queue metrics, node health, GPU utilization
Post-installation, we deliver job submission training, admin handoff, and complete documentation.
Slurm Consulting Services
If your Slurm installation isn’t performing as expected, or you need capacity planning:
- Current state audit — detailed review of slurm.conf, cgroup.conf, gres.conf, partition and QoS policies
- Performance assessment — job wait times, resource utilization rates, bottleneck identification
- Optimization recommendations — custom improvement report with implementation roadmap
- PBS/Torque to Slurm migration — script conversion, user migration, parallel run planning
- Capacity planning — scaling recommendations based on usage trends
Consulting is offered hourly or as fixed-scope projects. No long-term commitment required.
Managed Slurm Support
For continuous healthy operation with 24/7 coverage:
| Service | Scope |
|---|---|
| Proactive Monitoring | Queue depth, node state, GPU utilization — automatic intervention on threshold breach |
| Software Updates | Slurm version upgrades, security patches, plugin updates |
| User Management | Account creation, quota updates, partition modifications |
| Performance Optimization | Monthly performance report, bottleneck detection, improvement recommendations |
| Critical Response | 4-hour guaranteed response on system outages |
| Monthly Reporting | Uptime, queue statistics, resource utilization trends |
Why Mevasis for Slurm?
| Criterion | Mevasis | Alternatives |
|---|---|---|
| Experience | 12 years HPC, 40,000+ cores | General IT firms |
| Specialization | HPC and scientific computing only | Multi-sector, surface-level |
| Approach | Workload-specific policy design | Standard install |
| Support | 24/7, local language, same team | Call center, English-only |
| Independence | Not a hardware reseller | Vendor-biased recommendations |
Case Studies
Turbomachinery CFD Cluster
384-core on-prem HPC cluster with Slurm, Lustre, and RoCE — end-to-end deployment and operations.
Read case study →Genomics Analysis Pipeline
GPU-accelerated variant calling with Nextflow + Slurm on NVMe parallel storage — 10× speedup.
Read case study →Schedule a free discovery call to discuss your project. We’ll evaluate your current infrastructure and deliver a custom proposal within 2 business days.
Frequently Asked Questions
What do your Slurm services cover?
We offer four main service categories: installation and configuration, consulting and optimization, 24/7 monitoring and technical support, and training. We can deploy Slurm from scratch on new clusters, recover broken installations, or optimize existing configurations. All services are SLA-backed and contract-based.
Can I get consulting-only for my existing Slurm setup?
Absolutely. We offer hourly consulting or fixed-scope project engagements. We'll review your slurm.conf, partition structure, cgroup settings, and GPU GRES configuration, then deliver a customized optimization report.
What is the typical cost for Slurm installation?
Cost varies by cluster size and integration scope. Basic setup (1 controller + 16 nodes) has a fixed project fee; larger deployments use daily/monthly pricing. We prepare a custom quote during your free discovery call.
Do you offer Slurm training?
Yes. Two levels: System Administrator Training (2 days — installation, configuration, troubleshooting) and End User Training (1 day — job submission, scripting, GPU jobs). All training is hands-on, delivered on your infrastructure.
Can you help migrate from PBS/Torque to Slurm?
This is one of our most common engagements. We analyze your existing job scripts, convert them to Slurm format, migrate user accounts and group structures, and adapt your queue policies. We plan parallel run windows for zero-downtime transition.