storage-class
This component is responsible for provisioning StorageClasses
in an EKS cluster. See the list of guides and references
linked at the bottom of this README for more information.
A StorageClass provides part of the configuration for a PersistentVolumeClaim, which copies the configuration when it is created. Thus, you can delete a StorageClass without affecting existing PersistentVolumeClaims, and changes to a StorageClass do not propagate to existing PersistentVolumeClaims.
Usage
Stack Level: Regional, per cluster
This component can create storage classes backed by EBS or EFS, and is intended to be used with the corresponding EKS
add-ons aws-ebs-csi-driver
and aws-efs-csi-driver
respectively. In the case of EFS, this component also requires
that you have provisioned an EFS filesystem in the same region as your cluster, and expects you have used the efs
(previously eks/efs
) component to do so. The EFS storage classes will get the file system ID from the EFS component's
output.
Note: Default Storage Class
Exactly one StorageClass can be designated as the default StorageClass for a cluster. This default StorageClass is then used by PersistentVolumeClaims that do not specify a storage class.
Prior to Kubernetes 1.26, if more than one StorageClass is marked as default, a PersistentVolumeClaim without
storageClassName
explicitly specified cannot be created. In Kubernetes 1.26 and later, if more than one StorageClass
is marked as default, the last one created will be used, which means you can get by with just ignoring the default "gp2"
StorageClass that EKS creates for you.
EKS always creates a default storage class for the cluster, typically an EBS backed class named gp2
. Find out what the
default storage class is for your cluster by running this command:
# You only need to run `set-cluster` when you are changing target clusters
set-cluster <cluster-name> admin # replace admin with other role name if desired
kubectl get storageclass
This will list the available storage classes, with the default one marked with (default)
next to its name.
If you want to change the default, you can unset the existing default manually, like this:
SC_NAME=gp2 # Replace with the name of the storage class you want to unset as default
# You only need to run `set-cluster` when you are changing target clusters
set-cluster <cluster-name> admin # replace admin with other role name if desired
kubectl patch storageclass $SC_NAME -p '{"metadata": {"annotations":{"storageclass.kubernetes.io/is-default-class":"false"}}}'
Or you can import the existing default storage class into Terraform and manage or delete it entirely, like this:
SC_NAME=gp2 # Replace with the name of the storage class you want to unset as default
atmos terraform import eks/storage-class 'kubernetes_storage_class_v1.ebs["'${SC_NAME}'"]' $SC_NAME -s=core-usw2-dev
View the parameters of a storage class by running this command:
SC_NAME=gp2 # Replace with the name of the storage class you want to view
# You only need to run `set-cluster` when you are changing target clusters
set-cluster <cluster-name> admin # replace admin with other role name if desired
kubectl get storageclass $SC_NAME -o yaml
You can then match that configuration, except that you cannot omit allow_volume_exansion
.
ebs_storage_classes:
gp2:
make_default_storage_class: true
include_tags: false
# Preserve values originally set by eks/cluster.
# Set to "" to omit.
provisioner: kubernetes.io/aws-ebs
parameters:
type: gp2
encrypted: ""
Here's an example snippet for how to use this component.
eks/storage-class:
vars:
ebs_storage_classes:
gp2:
make_default_storage_class: false
include_tags: false
# Preserve values originally set by eks/cluster.
# Set to "" to omit.
provisioner: kubernetes.io/aws-ebs
parameters:
type: gp2
encrypted: ""
gp3:
make_default_storage_class: true
parameters:
type: gp3
efs_storage_classes:
efs-sc:
make_default_storage_class: false
efs_component_name: "efs" # Replace with the name of the EFS component, previously "eks/efs"
Variables
Required Variables
region
(string
) requiredAWS Region.
Optional Variables
ebs_storage_classes
optionalA map of storage class name to EBS parameters to create
Type:
map(object({
enabled = optional(bool, true)
make_default_storage_class = optional(bool, false)
include_tags = optional(bool, true) # If true, StorageClass will set our tags on created EBS volumes
labels = optional(map(string), null)
reclaim_policy = optional(string, "Delete")
volume_binding_mode = optional(string, "WaitForFirstConsumer")
mount_options = optional(list(string), null)
# Allowed topologies are poorly documented, and poorly implemented.
# According to the API spec https://kubernetes.io/docs/reference/generated/kubernetes-api/v1.26/#storageclass-v1-storage-k8s-io
# it should be a list of objects with a `matchLabelExpressions` key, which is a list of objects with `key` and `values` keys.
# However, the Terraform resource only allows a single object in a matchLabelExpressions block, not a list,
# the EBS driver appears to only allow a single matchLabelExpressions block, and it is entirely unclear
# what should happen if either of the lists has more than one element.
# So we simplify it here to be singletons, not lists, and allow for a future change to the resource to support lists,
# and a future replacement for this flattened object which can maintain backward compatibility.
allowed_topologies_match_label_expressions = optional(object({
key = optional(string, "topology.ebs.csi.aws.com/zone")
values = list(string)
}), null)
allow_volume_expansion = optional(bool, true)
# parameters, see https://github.com/kubernetes-sigs/aws-ebs-csi-driver/blob/master/docs/parameters.md
parameters = object({
fstype = optional(string, "ext4") # "csi.storage.k8s.io/fstype"
type = optional(string, "gp3")
iopsPerGB = optional(string, null)
allowAutoIOPSPerGBIncrease = optional(string, null) # "true" or "false"
iops = optional(string, null)
throughput = optional(string, null)
encrypted = optional(string, "true")
kmsKeyId = optional(string, null) # ARN of the KMS key to use for encryption. If not specified, the default key is used.
blockExpress = optional(string, null) # "true" or "false"
blockSize = optional(string, null)
})
provisioner = optional(string, "ebs.csi.aws.com")
# TODO: support tags
# https://github.com/kubernetes-sigs/aws-ebs-csi-driver/blob/master/docs/tagging.md
}))Default value:
{ }
efs_storage_classes
optionalA map of storage class name to EFS parameters to create
Type:
map(object({
enabled = optional(bool, true)
make_default_storage_class = optional(bool, false)
labels = optional(map(string), null)
efs_component_name = optional(string, "eks/efs")
reclaim_policy = optional(string, "Delete")
volume_binding_mode = optional(string, "Immediate")
# Mount options are poorly documented.
# TLS is now the default and need not be specified. https://github.com/kubernetes-sigs/aws-efs-csi-driver/tree/master/docs#encryption-in-transit
# Other options include `lookupcache` and `iam`.
mount_options = optional(list(string), null)
parameters = optional(object({
basePath = optional(string, "/efs_controller")
directoryPerms = optional(string, "700")
provisioningMode = optional(string, "efs-ap")
gidRangeStart = optional(string, null)
gidRangeEnd = optional(string, null)
# Support for cross-account EFS mounts
# See https://github.com/kubernetes-sigs/aws-efs-csi-driver/tree/master/examples/kubernetes/cross_account_mount
# and for gritty details on secrets: https://kubernetes-csi.github.io/docs/secrets-and-credentials-storage-class.html
az = optional(string, null)
provisioner-secret-name = optional(string, null) # "csi.storage.k8s.io/provisioner-secret-name"
provisioner-secret-namespace = optional(string, null) # "csi.storage.k8s.io/provisioner-secret-namespace"
}), {})
provisioner = optional(string, "efs.csi.aws.com")
}))Default value:
{ }
eks_component_name
(string
) optionalThe name of the EKS component for the cluster in which to create the storage classes
Default value:
"eks/cluster"
helm_manifest_experiment_enabled
(bool
) optionalEnable storing of the rendered manifest for helm_release so the full diff of what is changing can been seen in the plan
Default value:
false
kube_data_auth_enabled
(bool
) optionalIf
true
, use anaws_eks_cluster_auth
data source to authenticate to the EKS cluster.
Disabled bykubeconfig_file_enabled
orkube_exec_auth_enabled
.Default value:
false
kube_exec_auth_aws_profile
(string
) optionalThe AWS config profile for
aws eks get-token
to useDefault value:
""
kube_exec_auth_aws_profile_enabled
(bool
) optionalIf
true
, passkube_exec_auth_aws_profile
as theprofile
toaws eks get-token
Default value:
false
kube_exec_auth_enabled
(bool
) optionalIf
true
, use the Kubernetes providerexec
feature to executeaws eks get-token
to authenticate to the EKS cluster.
Disabled bykubeconfig_file_enabled
, overrideskube_data_auth_enabled
.Default value:
true
kube_exec_auth_role_arn
(string
) optionalThe role ARN for
aws eks get-token
to useDefault value:
""
kube_exec_auth_role_arn_enabled
(bool
) optionalIf
true
, passkube_exec_auth_role_arn
as the role ARN toaws eks get-token
Default value:
true
kubeconfig_context
(string
) optionalContext to choose from the Kubernetes config file.
If supplied,kubeconfig_context_format
will be ignored.Default value:
""
kubeconfig_context_format
(string
) optionalA format string to use for creating the
kubectl
context name when
kubeconfig_file_enabled
istrue
andkubeconfig_context
is not supplied.
Must include a single%s
which will be replaced with the cluster name.Default value:
""
kubeconfig_exec_auth_api_version
(string
) optionalThe Kubernetes API version of the credentials returned by the
exec
auth pluginDefault value:
"client.authentication.k8s.io/v1beta1"
kubeconfig_file
(string
) optionalThe Kubernetes provider
config_path
setting to use whenkubeconfig_file_enabled
istrue
Default value:
""
kubeconfig_file_enabled
(bool
) optionalIf
true
, configure the Kubernetes provider withkubeconfig_file
and use that kubeconfig file for authenticating to the EKS clusterDefault value:
false
Context Variables
The following variables are defined in the context.tf
file of this module and part of the terraform-null-label pattern.
context.tf
file of this module and part of the terraform-null-label pattern.additional_tag_map
(map(string)
) optionalAdditional key-value pairs to add to each map in
tags_as_list_of_maps
. Not added totags
orid
.
This is for some rare cases where resources want additional configuration of tags
and therefore take a list of maps with tag key, value, and additional configuration.Required: No
Default value:
{ }
attributes
(list(string)
) optionalID element. Additional attributes (e.g.
workers
orcluster
) to add toid
,
in the order they appear in the list. New attributes are appended to the
end of the list. The elements of the list are joined by thedelimiter
and treated as a single ID element.Required: No
Default value:
[ ]
context
(any
) optionalSingle object for setting entire context at once.
See description of individual variables for details.
Leave string and numeric variables asnull
to use default value.
Individual variable settings (non-null) override settings in context object,
except for attributes, tags, and additional_tag_map, which are merged.Required: No
Default value:
{
"additional_tag_map": {},
"attributes": [],
"delimiter": null,
"descriptor_formats": {},
"enabled": true,
"environment": null,
"id_length_limit": null,
"label_key_case": null,
"label_order": [],
"label_value_case": null,
"labels_as_tags": [
"unset"
],
"name": null,
"namespace": null,
"regex_replace_chars": null,
"stage": null,
"tags": {},
"tenant": null
}delimiter
(string
) optionalDelimiter to be used between ID elements.
Defaults to-
(hyphen). Set to""
to use no delimiter at all.Required: No
Default value:
null
descriptor_formats
(any
) optionalDescribe additional descriptors to be output in the
descriptors
output map.
Map of maps. Keys are names of descriptors. Values are maps of the form
\{<br/> format = string<br/> labels = list(string)<br/> \}
(Type isany
so the map values can later be enhanced to provide additional options.)
format
is a Terraform format string to be passed to theformat()
function.
labels
is a list of labels, in order, to pass toformat()
function.
Label values will be normalized before being passed toformat()
so they will be
identical to how they appear inid
.
Default is{}
(descriptors
output will be empty).Required: No
Default value:
{ }
enabled
(bool
) optionalSet to false to prevent the module from creating any resources
Required: NoDefault value:
null
environment
(string
) optionalID element. Usually used for region e.g. 'uw2', 'us-west-2', OR role 'prod', 'staging', 'dev', 'UAT'
Required: NoDefault value:
null
id_length_limit
(number
) optionalLimit
id
to this many characters (minimum 6).
Set to0
for unlimited length.
Set tonull
for keep the existing setting, which defaults to0
.
Does not affectid_full
.Required: No
Default value:
null
label_key_case
(string
) optionalControls the letter case of the
tags
keys (label names) for tags generated by this module.
Does not affect keys of tags passed in via thetags
input.
Possible values:lower
,title
,upper
.
Default value:title
.Required: No
Default value:
null
label_order
(list(string)
) optionalThe order in which the labels (ID elements) appear in the
id
.
Defaults to ["namespace", "environment", "stage", "name", "attributes"].
You can omit any of the 6 labels ("tenant" is the 6th), but at least one must be present.Required: No
Default value:
null
label_value_case
(string
) optionalControls the letter case of ID elements (labels) as included in
id
,
set as tag values, and output by this module individually.
Does not affect values of tags passed in via thetags
input.
Possible values:lower
,title
,upper
andnone
(no transformation).
Set this totitle
and setdelimiter
to""
to yield Pascal Case IDs.
Default value:lower
.Required: No
Default value:
null
labels_as_tags
(set(string)
) optionalSet of labels (ID elements) to include as tags in the
tags
output.
Default is to include all labels.
Tags with empty values will not be included in thetags
output.
Set to[]
to suppress all generated tags.
Notes:
The value of thename
tag, if included, will be theid
, not thename
.
Unlike othernull-label
inputs, the initial setting oflabels_as_tags
cannot be
changed in later chained modules. Attempts to change it will be silently ignored.Required: No
Default value:
[
"default"
]name
(string
) optionalID element. Usually the component or solution name, e.g. 'app' or 'jenkins'.
This is the only ID element not also included as atag
.
The "name" tag is set to the fullid
string. There is no tag with the value of thename
input.Required: No
Default value:
null
namespace
(string
) optionalID element. Usually an abbreviation of your organization name, e.g. 'eg' or 'cp', to help ensure generated IDs are globally unique
Required: NoDefault value:
null
regex_replace_chars
(string
) optionalTerraform regular expression (regex) string.
Characters matching the regex will be removed from the ID elements.
If not set,"/[^a-zA-Z0-9-]/"
is used to remove all characters other than hyphens, letters and digits.Required: No
Default value:
null
stage
(string
) optionalID element. Usually used to indicate role, e.g. 'prod', 'staging', 'source', 'build', 'test', 'deploy', 'release'
Required: NoDefault value:
null
tags
(map(string)
) optionalAdditional tags (e.g.
{'BusinessUnit': 'XYZ'}
).
Neither the tag keys nor the tag values will be modified by this module.Required: No
Default value:
{ }
tenant
(string
) optionalID element (Rarely used, not included by default). A customer identifier, indicating who this instance of a resource is for
Required: NoDefault value:
null
Outputs
storage_classes
Storage classes created by this module
Dependencies
Requirements
terraform
, version:>= 1.3.0
aws
, version:>= 4.9.0
helm
, version:>= 2.0
kubernetes
, version:>= 2.22.0
Providers
aws
, version:>= 4.9.0
kubernetes
, version:>= 2.22.0
Modules
Name | Version | Source | Description |
---|---|---|---|
efs | 1.5.0 | cloudposse/stack-config/yaml//modules/remote-state | n/a |
eks | 1.5.0 | cloudposse/stack-config/yaml//modules/remote-state | n/a |
iam_roles | latest | ../../account-map/modules/iam-roles | n/a |
this | 0.25.0 | cloudposse/label/null | n/a |
Resources
The following resources are used by this module:
kubernetes_storage_class_v1.ebs
(resource)kubernetes_storage_class_v1.efs
(resource)
Data Sources
The following data sources are used by this module:
aws_eks_cluster_auth.eks
(data source)
Related How-to Guides
- EBS CSI Migration FAQ
- Migrating Clusters From gp2 to gp3 EBS Volumes
- Kubernetes: Change the Default StorageClass