---
description: "Practical notes before running a Verda Instant Cluster: quotas and limits, networking behaviour, node lifecycle, and the common pitfalls worth avoiding."
revision_date: 04.08.2026
---

# Good to know

This page has been reorganized. Its content now lives here:

* Cluster architecture, node naming, storage and preinstalled software: [Architecture](https://docs.verda.com/clusters/instant-clusters/basics/architecture/)
* Default open ports, external SSH to worker nodes, InfiniBand partitioning: [Networking and ports](https://docs.verda.com/clusters/instant-clusters/basics/networking/)
* Changing the Slurm configuration on Slinky: [Slinky → Overview & architecture](https://docs.verda.com/clusters/instant-clusters/slinky/#changing-the-slurm-configuration)
* NVIDIA userspace inside Slinky pods: [Slinky → Shared jail](https://docs.verda.com/clusters/instant-clusters/slinky/shared-jail/#where-the-nvidia-userspace-comes-from)
* Seamless SSH for cluster users: [Slinky → User management](https://docs.verda.com/clusters/instant-clusters/slinky/users/#how-users-log-in)