K2cloud Versionless Space Profile Naming Guide

This guide explains how to read the names of K2cloud’s versionless Space Profiles. These profiles standardize sizing across environments and are available out-of-the-box (OOTB).

• Existing customers: ask us to enable these profiles on your tenant.
• New customers: profiles are available at provisioning.

How to read a profile name

Pattern

<Product>-c<CPUs>-m<MemoryGB>[-r<Replicas> | -r<Min>-<Max>][-pg | -mngd | -noSdB]

Tokens

• Product — Studio or Fabric.
• c# — vCPU per pod (e.g., c4 = 4 vCPU).
• m# — memory (GB) per pod (e.g., m32 = 32 GB).
• r# — fixed replica count for Fabric (Studio is single-pod).
• r<Min>-<Max> — autoscale replica range for Fabric (min/max).
• -pg — Studio only: includes an embedded PostgreSQL system DB.
• -mngd — Fabric only: uses a managed PostgreSQL service + object storage bucket.
• -noSdB — Fabric only: no system DB is provisioned by K2cloud.

Storage classes are cloud-aware and set via your Foundation (e.g., GCP regional-pd, Azure managed, AWS gp3).

Examples

• Studio-c4-m32 → Studio, 4 vCPU / 32 GB, 1 pod, SQLite system DB.
• Studio-c4-m32-pg → Studio, 4 vCPU / 32 GB, 1 pod, embedded PostgreSQL system DB.
• Fabric-c8-m64-r3-noSdB → Fabric, 8 vCPU / 64 GB per pod, 3 replicas, no system DB.
• Fabric-c4-m32-r5-mngd → Fabric, 4 vCPU / 32 GB per pod, 5 replicas, managed PG + bucket.
• Fabric-c2-m16-r2-4-mngd → Fabric, 2 vCPU / 16 GB, autoscale 2–4 replicas, managed PG + bucket.

Catalog of current profiles

Studio

Fabric (no system DB)

Fabric (managed Postgres + bucket)

Autoscale naming (Fabric)

When autoscaling is enabled, the replica token shows the minimum and maximum replicas:

• Format: -r<Min>-<Max>
• Example: Fabric-c2-m16-r2-4-mngd → min 2, max 4 replicas.

Choosing a profile (quick pointers)

• Studio: start with c2/m16 or c4/m32; use -pg if you prefer a PostgreSQL system DB over SQLite.
• Fabric: use -mngd when you want cloud-managed PostgreSQL + object storage; choose -noSdB when your space will not provision a system DB.
• Use higher c/m or more replicas for heavier workloads or stricter HA.

If you’d like these profiles enabled on your tenant—or need autoscale variants—please open a ticket with the profile names you want.