Quotas

A quota is a limit on how much of a resource you can use. Each tenant has a limit per resource, such as instances, vCPUs or block storage. A create that would take a tenant over a limit is refused with a quota error, and nothing is created.

This page covers:

• how tenant and organization limits work together;
• the resources that have a quota;
• default limits;
• viewing your quotas in the portal, with fm and in Terraform;
• requesting an increase.

Tenant and organization limits

Every tenant has its own limit for each resource. Your organization also has a pool per resource: an organization-level limit across all of its tenants. Each pool works in one of two ways:

• Overcommit. The organization limit caps the total usage of all tenants together. The tenants' limits may add up to more than the organization limit, so a tenant can use more while the others use less.
• Partition. The organization limit is divided between the tenants. Their limits may not add up to more than the organization limit.

A create must fit both its tenant's limit and the organization pool.

Resources

The portal groups the resources as below. The key is the resource's resourceType, which the fm CLI and the Terraform provider use.

Storage limits are in GB, as the platform reports them. Memory (ram) is counted in MB and shown in GB, where 1 GB = 1024 MB. Block storage follows the volume sizes described on the Block Volumes page.

A limit of 0 means no allowance: you cannot create that resource. It never means unlimited.

When usage is shown as —, it is unknown right now, for example because it could not be read. It does not mean zero. Your limit still applies.

Compute

Resource	Key	Unit	Counts
Instances	instance	count	Instances
vCPUs	vcpu	count	vCPUs across your instances
Memory	ram	MB	Memory across your instances, shown as GB
Web servers	webserver	count	Managed web servers
Web server storage	webserver_storage	GB	Storage of your managed web servers

Custom images

Resource	Key	Unit	Counts
Images	image	count	Custom images you have uploaded
Image storage	image_storage	GB	Storage used by your custom images

These two limits are enforced by the compute service and cannot be raised with a quota request. To raise them, contact support.

The portal and fm show these rows only when your tenant has an image allowance. Terraform always returns the image and image_storage keys, with a limit of 0 when there is no allowance.

Storage

Resource	Key	Unit	Counts
Block storage	volume	GB	Total size of your volumes
Block volumes	volume_count	count	Number of volumes
Snapshots	snapshot	count	Snapshots

Network

Resource	Key	Unit	Counts
Networks	network	count	Networks
Subnets	subnet	count	Subnets
Public IPs	public_ip	count	Public IPs
Load balancers	load_balancer	count	Load balancers. Listeners and pools are not counted.
Application gateways	app_gateway	count	Application gateways
Security groups	security_group	count	Security groups you create. Groups the platform creates for managed services, Kubernetes or application gateways are not counted.

Managed databases

Resource	Key	Unit	Counts
Databases	database	count	Managed databases
Database storage	database_storage	GB	Storage of your databases

Managed cache

Resource	Key	Unit	Counts
Cache instances	cache	count	Managed cache instances
Cache memory	cache_memory	GB	Memory of your caches
Cache storage	cache_storage	GB	Storage of your caches

Managed messaging

Resource	Key	Unit	Counts
Message brokers	messaging	count	Managed message brokers
Message broker storage	messaging_storage	GB	Storage of your message brokers

Object storage

Resource	Key	Unit	Counts
Object storage	object_storage	GB	Object storage in your buckets
Buckets	buckets	count	Buckets
S3 credentials	s3_credentials	count	S3 credentials

Container registry

Resource	Key	Unit	Counts
Registry storage	registry_storage	GB	One allowance across your registry and its pull-through caches together

Kubernetes

Resource	Key	Unit	Counts
Kubernetes clusters	kubernetes_cluster	count	Managed Kubernetes clusters
Kubernetes nodes	kubernetes_node	count	Worker nodes across your clusters

Kontor

Resource	Key	Unit	Counts
Kontor users	kontor_user	count	Kontor users
Kontor mail storage	kontor_mail_storage	GB	Mail storage
Kontor file storage	kontor_file_storage	GB	File storage
Kontor domains	kontor_domain	count	Kontor domains

Default limits

A new tenant starts with the platform's default limits. Frostmoln sets the defaults and can change them, so they are not listed here. To see the limits that apply to you, view a tenant's quotas: a new tenant shows the defaults.

If the defaults are too small for something you plan to run, such as a production Kubernetes cluster with large worker nodes, request an increase before you create it.

View your quotas

Portal. Choose Quotas in the main navigation. The page shows the limits and usage of the current tenant, and the organization's pools with a per-tenant breakdown.

fm CLI. The commands act on the active tenant:

fm tenant quota
fm tenant quota -o json
fm tenant quota --org

fm tenant quota prints a table. -o json prints {"scope":"tenant","quotas":[…]}, where each quota has resourceType, label, category, used, limit and unit; unknown usage is null. With --org, the JSON also has orgId and tenants. --org shows the organization's pools, with an ALLOCATED column for how much of each pool is allocated to tenants. It needs a signed-in user; it does not work with an API key. See fm tenant.

Terraform. The frostmoln_tenant_quotas data source reads the quotas of the provider's tenant. quotas is keyed by resource key. used is null when usage is unknown, so check for that before you calculate with it:

data "frostmoln_tenant_quotas" "this" {}

locals {
  vcpu = data.frostmoln_tenant_quotas.this.quotas["vcpu"]
}

output "vcpu_left" {
  # Null when usage is unknown, rather than a number that may be wrong.
  value = local.vcpu.used == null ? null : local.vcpu.limit - local.vcpu.used
}

Request an increase

Portal. Choose Request increase on a quota row, or go to Support → Quota requests.

fm CLI.

fm tenant quota request vcpu 64 --reason "Production Kubernetes cluster"
fm tenant quota request volume 2000 --org --reason "Data warehouse"

The arguments are <resource-type> (the key) and <new-limit>. Give the new limit in the resource's own unit: MB for Memory, GB for storage, otherwise a count. Add --org to request a change to the organization pool instead of the tenant's limit.

• A tenant request is decided at once. It is approved if the new limit fits within what the organization pool allows, and denied with a reason if it does not. The portal shows the largest value that would be approved, and fm refuses a higher value before it sends the request.
• An organization request goes to Frostmoln for review and stays pending until it is decided.
• A lower tenant limit is applied at once, even below what the tenant uses now. After that, new creates of that resource are refused. fm tenant quota request refuses a lower value unless you pass --yes.
• You can have only one pending request per resource at a time.
• --org needs a signed-in user, both to view and to request. It does not work with an API key.
• With an API key, a tenant request (and cancelling one) needs the quotas:write scope; compute:write is not enough. A workload identity cannot request quota changes. See API keys.

Custom image limits cannot be requested this way; contact support.

Related

• fm CLI Reference: fm tenant
• Terraform Reference
• Billing