diff --git a/modules/instances/README.md b/modules/instances/README.md index d3d8542..2f9d52f 100644 --- a/modules/instances/README.md +++ b/modules/instances/README.md @@ -1,4 +1,5 @@ - [Instances](#instances) + - [Using Flex Shapes](#using-flex-shapes) - [Assign Public IP to an instnace](#assign-public-ip-to-an-instnace) - [Limitations](#limitations) - [Examples](#examples) @@ -10,6 +11,13 @@ Each object in the input list represent an instance that can have its own config Using this module you can attach multiple VNICs to your instance. Moreover, you can assign secondary IP (e.g as floating IP) to each of VNIC. +## Using Flex Shapes +You have the options to use AMD/Intel flex shapes. Just set `flex_shape_config` key in `var.instances.*.config` varible per instance. The variable `flex_shape_config` should have the following two keys `ocpus` and `memory_in_gbs`. See example below. + +Note that, Flex Shape works only with AMD shape `VM.Standard.E3.Flex` and `VM.Standard.E4.Flex`, and Intel `VM.Standard3.Flex`. + +Leave `flex_shape_config` empty `{}` if it is not needed. + # Assign Public IP to an instnace To attach public IP to any private IP created in this module, you have to do that in `public_ip` module. Refer to `public_ip` module for how to attach private IP to public IP. Using this module output, you can get the private ip OCID and use it in `public_ip` module. @@ -31,6 +39,10 @@ locals { state = "RUNNING" config = { shape = "ocixxxxxx.xxxxxx.xxxxx" + flex_shape_config = { + ocpus = 8 + memory_in_gbs = 32 + } image_id = "ocixxxxxx.xxxxxx.xxxxx" subnet = { id = "ocixxxxxx.xxxxxx.xxxxx" @@ -57,6 +69,7 @@ locals { state = "RUNNING" config = { shape = "ocixxxxxx.xxxxxx.xxxxx" + flex_shape_config = {} image_id = "ocixxxxxx.xxxxxx.xxxxx" subnet = { id = "ocixxxxxx.xxxxxx.xxxxx" diff --git a/modules/kubernetes/README.md b/modules/kubernetes/README.md index 32b13cb..f5be027 100644 --- a/modules/kubernetes/README.md +++ b/modules/kubernetes/README.md @@ -10,7 +10,7 @@ When the node pool configuration is updated, the existing worker nodes will keep ## Using Flex Shapes You have the options to use AMD flex shaeps. Just set `flex_shape_config` key in `node_pools` varible per nood pool. The variable `flex_shape_config` should have the following two keys `ocpus` and `memory_in_gbs`. See example below. -Note that, Flex Shape works only with AMD shape `VM.Standard.E3.Flex` and `VM.Standard.E4.Flex`. +Note that, Flex Shape works only with AMD shape `VM.Standard.E3.Flex` and `VM.Standard.E4.Flex`, and Intel `VM.Standard3.Flex`. ## Example Creating a cluster with 2 node pools diff --git a/releases.md b/releases.md index ec4f5b8..db235db 100644 --- a/releases.md +++ b/releases.md @@ -1,3 +1,30 @@ +# v2.1.0: +_**Please see breaking changes section before upgrading.**_ +## **New** +* `instances`: add posibility to use flex shape configuration + +## _**Breaking Changes**_ +* `instances` modules input is updated. A new key `flex_shape_config` is now required under `var.instances.*.config`. + * Add `flex_shape_config` and set its value to `{}`. Example of partial instance object. See module's readme for full example. +``` +"instance-a" = { + name = "instance-a" + availability_domain_name = "ocixxxxxx.xxxxxx.xxxxx" + fault_domain_name = "ocixxxxxx.xxxxxx.xxxxx" + compartment_id = "ocixxxxxx.xxxxxx.xxxxx" + volume_size = 500 + autherized_keys = "ssh-rsa xxxxxxxxxxxxxxxxxxxxxx\n ssh-rsa xxxxxxxxxxxxxxxxxxxxxx" + state = "RUNNING" + config = { + shape = "ocixxxxxx.xxxxxx.xxxxx" + flex_shape_config = {} <-------------------------------------------------------------------- note this line + ... + ... + ... + ... + } +} +``` # v2.0.1: ## Fix * `instances`: fix output of module (use `instance.id` instead of `k.id`)