Support structured config #575

ljtfreitas · 2024-04-25T17:09:14Z

Context

Currently structured configs are not supported properly by Pulumi k8s operator.

As Pulumi programs can be configured using YAML files, it looks fair to be able to configure a structured, complex config using the Stack CRD as well.

Proposed changes

The proposed approach is to introduce a new field called configRefs, rather than change the current config one.

configRefs has a format close to other fields like envRefs and secretsRefs, supporting the same env and filesystem options as well (secret is not supported because just use secretsRefs looks better), and two new options:

ConfigLiteralRef
ConfigMapRef

ConfigLiteralRef

ConfigLiteralRef refers to both a simple or possibly structured, nested value in YAML format, in the same way that people would put in regular Pulumi.<stack>.yaml files:

configRefs:
   stack-config:
     type: Literal
     value:
        my-config:
          my-nested-field:
             field: "value"

The value content will be handled as a map of any type of value; so just simple key-pairs in YAML format are supported as well:

configRefs:
   stack-config:
     type: Literal
     value:
       field-a: value-a
       field-b: value-b

ConfigMapRef

ConfigMapRef refers to a Kubernetes ConfigMap. The implementation assumes that the content from the ConfigMap's key is a structured, nested content in YAML format. The idea is to enable people to just put their configs from Pulumi.<stack>.yaml files in a ConfigMap, and the operator will be able to handle it.

So given a ConfigMap called my-stack-config, like that:

apiVersion: v1
kind: ConfigMap
metadata:
  name: my-stack-config
data:
  Pulumi.my-stack.yaml: |
    my-config:
      my-nested-field:
          field: "value"

It could be consumed as a config to a Pulumi stack like that:

configRefs:
   stack-config:
     type: ConfigMap
     configmap:
       name: my-stack-config
       key: Pulumi.my-stack.yaml

Configs merge strategy

All configs resolved from configRefs fields will be merged with the ones from the config field. configRefs resolved key-values will override config ones with the same key.

Configs `set` strategy

All configs will be set using the path flag. So the config field will start to support nested values as well.

It's especially useful for some complex fields from providers. An example:

config:
   aws:assumeRole.roleArn: arn/my-aws-arn

Passing the value above with the --path flag, it will be configured like that in the result stack config:

aws:assumeRole:
  roleArn: arn/my-aws-arn

And simple key-value pairs keep working also.

Related issues (optional)

#258
#516

github-actions · 2024-04-25T17:09:28Z

PR is now waiting for a maintainer to run the acceptance tests. This PR will only perform build and linting.
Note for the maintainer: To run the acceptance tests, please comment /run-acceptance-tests on the PR

rquitales · 2024-04-25T20:52:52Z

/run-acceptance-tests

github-actions · 2024-04-25T20:53:15Z

Please view the PR build - https://github.com/pulumi/pulumi-kubernetes-operator/actions/runs/8839237664

ljtfreitas · 2024-04-25T21:08:46Z

cc @rquitales @EronWright

blampe · 2024-04-25T21:56:40Z

This is awesome, thank you for the contribution! One initial thought -- still need to take a deeper look.

The idea is to enable people to just put their configs from Pulumi..yaml files in a ConfigMap, and the operator will be able to handle it.

This is a great idea. What do you think about using Pulumi's native marshalers? That would give you a little validation on top, and it might also make your flattening logic unnecessary (although I'm not 100%). The encoding package should do the trick, and you would unmarshal into a ProjectStack.

ljtfreitas · 2024-04-26T16:37:48Z

@blampe Thanks. About your suggestion maybe I haven't expressed myself properly. My initial idea, and that's the implementation is actually doing, is to put in a ConfigMap just the config node from a Pulumi.<stack>.yaml file, not the whole ProjectStack. Pulumi.<stack>.yaml can contain fields like secretsprovider, as the struct that you mentioned have, but it was not my idea to be able to configure them through a ConfigMap.

However, it looks like a good idea to me 🤔 ; so we could move the whole stack file (Pulumi.<stack>.yaml) to a ConfigMap and we even wouldn't have to configure things like secretsprovider or backend in the CRD 🤔 ; we could just unmarshal the ConfigMap key's content as a ProjectStack. Not sure if it's related with the specific topic of structured config but looks good to me.

blampe · 2024-04-26T18:10:19Z

@blampe Thanks. About your suggestion maybe I haven't expressed myself properly. My initial idea, and that's the implementation is actually doing, is to put in a ConfigMap just the config node from a Pulumi.<stack>.yaml file, not the whole ProjectStack. Pulumi.<stack>.yaml can contain fields like secretsprovider, as the struct that you mentioned have, but it was not my idea to be able to configure them through a ConfigMap.

However, it looks like a good idea to me 🤔 ; so we could move the whole stack file (Pulumi.<stack>.yaml) to a ConfigMap and we even wouldn't have to configure things like secretsprovider or backend in the CRD 🤔 ; we could just unmarshal the ConfigMap key's content as a ProjectStack. Not sure if it's related with the specific topic of structured config but looks good to me.

Ah, I had misunderstood what all was in the ConfigMap, I didn't realize it was just the config key. In that case you could marshal the config.Map instead of the whole ProjectStack as I had suggested earlier. I didn't mean to increase the scope of your change!

ljtfreitas · 2024-04-27T21:43:09Z

@blampe Don't worry, maybe I haven't expressed myself properly in the PR description (English is not my first language so sometimes talking about complex stuff is just...complex 😅). About your suggestion, I'm trying to refactor the code to use config.Map but I'm struggling to read the content. When I try to unmarshal I see this error:

could not parse <my-config> as a configuration key (configuration keys should be of the form `<namespace>:<name>`

Looking at the code that deserializes a YAML to a ProjectStack instance (in the Automation API code), config keys are namespaced (with the project name) before the config.Map's unmarshal 😕 ...and I can't use the same function because it's private. So config keys must be namespaced in the ConfigMap.

Not sure if we need to enforce that here because this behavior will be applied anyway during the config-set phase (and speaking for myself I don't namespace my configs because of this)

ljtfreitas · 2024-05-08T17:18:35Z

hey friends, any news here? we really need this feature

@blampe @rquitales

blampe · 2024-05-09T00:17:24Z

Sorry for the delay @ljtfreitas, I plan on playing around with this tomorrow!

blampe

Apologies again for the delay. I was able to spend some hands-on time with this and have a much better feel for things.

The biggest issue I see is that I think we'll want ConfigRefs to be an array. Let me know if you think that sounds reasonable.

I'm still playing around with some of the config merging logic locally, in case there's some room to simplify things.

blampe · 2024-05-10T21:08:10Z

test/testdata/structured-config-refs/go.mod

@@ -0,0 +1,90 @@
+module structured-config-refs


Out of curiosity, did you include a Go program here because you're using Go for your own projects?

I'm a big fan of YAML projects for tests, but this is fine.

I guess I was a bit biased here in fact 😅, I wrote the example/test program in Go because I wanted to be sure that we could deserialize a structured, complex config in the way that we usually do in Go

blampe · 2024-05-10T21:08:35Z

test/stack_controller_test.go

@@ -14,6 +15,7 @@ import (

 	"github.com/pulumi/pulumi-kubernetes-operator/pkg/apis/pulumi/shared"
 	pulumiv1 "github.com/pulumi/pulumi-kubernetes-operator/pkg/apis/pulumi/v1"
+	"sigs.k8s.io/yaml"


👍 We've had issues with marshaling when not using the k8s fork, so this is good.

blampe · 2024-05-10T22:45:10Z

deploy/crds/pulumi.com_stacks.yaml

+                description: (optional) ConfigRefs is the configuration for this stack,
+                  which can be specified through ConfigRef. If this is omitted, configuration
+                  is assumed to be checked in and taken from the source repository.
+                  If present, ConfigRefs values will be merged with values passed


We should clarify which one has precedence -- e.g. ConfigRefs will overwrite Configs.

but ConfigRefs will not overwrite Configs, right? in fact both fields will be merged

I guess the description could be improved to explain the ConfigRefs will overwrite Configs with the same key name

blampe · 2024-05-10T22:48:08Z

deploy/crds/pulumi.com_stacks.yaml

+                additionalProperties:
+                  description: ConfigRef identifies a resource from which config information
+                    can be loaded. Environment variables, files on the filesystem,
+                    Kubernetes Secrets, ConfigMaps, structured and config literal
+                    values strings are currently supported.
+                  properties:
+                    configmap:


ConfigRefs will need to be an array to allow supporting more than one ref with predictable precedence.

For example the user might want to specify more than one secret ref, and if each of those refs contains the same key, then the value from the last ref should win.

As a map, the user can't specify the same ref type more than once, and the map's iteration order isn't guaranteed.

Interesting. It could be an array, but do we really need this? 🤔 ConfigRefs would have the same semantics as EnvRefs field.

In EnvRefs, one could specify more than one secret, or more than one env var, with the same key name and the order will not be guaranteed; but this would look like a misconfiguration to me (using the same secret key twice), the behavior seems reasonable.

also, the ConfigRefs spec above supports to specify the same ref type more than once, like:

configRefs: first-secret: type: Secret secret: name: my-secret key: key-name second-secret: type: Secret secret: name: my-another-secret key: key-name

EnvRefs makes sense as a map because the keys are just variable names, and I agree you would never need to specify the same variable twice.

ConfigRefs is different because each ref can potentially be its own bag of key/values. All of those key/values need to be coalesced and we want that process to be well-defined.

also, the ConfigRefs spec above supports to specify the same ref type more than once, like:

Yes, you're right, I wasn't clear on the key when I wrote that. I'm still worried about the iteration order, though. Consider the case where you have some "base" config that you want individual stacks to be able to override:

configRefs: base: type: Structured value: my-config: base-value override: type: Structured value: my-config: override-value

I would expect the override value to always be applied, but that's not guaranteed.

Ok, it makes sense. I will change it 👍

Thinking about the best way to do it 🤔 ...maybe an "array of maps"?

configRefs: - base: type: Structured value: my-config: base-value - override: type: Structured value: my-config: override-value

Or an object with a "name" field (or "key" or something)?

configRefs: - name: base type: Structured value: my-config: base-value - name: override type: Structured value: my-config: override-value

I prefer the first one but any thoughts?

@blampe just added a new test case for this scenario

blampe · 2024-05-10T22:56:07Z

pkg/controller/stack/stack_controller.go

+	sort.Slice(configValues, func(i, j int) bool {
+		return configValues[i].Key < configValues[j].Key
+	})


I need to double check this, but I believe keys can be specified in flat or exploded form, so for example object.nums[0]: 1 and object: { nums: [1] }. I think you already handle this by flattening, but it would be worth a test to verify that 2 refs specifying the same key in different ways will be merged together.

Is the second example a valid syntax? As a json?

I just doubled checked -- these are two ways of describing the same thing:

stack:deepobject: nums: - 1 stack:flatobject: '{"nums": [1]}

It doesn't look like stack:object.nums[0]: 1 is valid.

I think it would be totally fine to say this is an edge case we don't handle, given that it's unlikely to come up.

blampe · 2024-05-10T22:57:41Z

test/stack_controller_test.go

@@ -205,6 +207,386 @@ var _ = Describe("Stack Controller", func() {
 		})
 	})

+	Context("configuring a stack using ConfigRefs", func() {


Thank you enormously for adding tests!

blampe · 2024-05-10T23:02:47Z

test/stack_controller_test.go

+					Stack:   stackName,
+					GitSource: &shared.GitSource{
+						ProjectRepo: baseDir,
+						RepoDir:     "test/testdata/config-refs",


Could we just test/testdata/structured-config-refs for all of these? This is testing we can read the configs, seems worth also confirming the structure was handled correctly.

Actually this test is trying to read a file, and in a similar way that I implemented with Literal, I used the same existing behavior to FileRef (the operator already use it to load envvars and secrets from files), which means that just loads a single value. So this test couldn't use test/testdata/structured-config-refs

Do you believe that would be worth implementing a file behavior specifically to load configs? So we could support structured values as well.

blampe · 2024-05-10T23:13:08Z

pkg/controller/stack/stack_controller.go

+				structuredConfig := StructuredConfig(configMapContent).Flatten()
+				allConfigs = append(allConfigs, structuredConfig...)
+			}
+		case shared.ConfigResourceSelectorStructured:


I'm a little confused about the need for both "structured" and "literal".

The legacy resolveResourceRef logic doesn't do any special handling for "literal", but we have an opportunity to be more flexible here.

In other words, it's odd to have two ways to express the same thing:

configRefs: stack-config: type: Literal value: foo -- configRefs: stack-config: type: Structured value: foo

But this is net-new:

configRefs: stack-config: type: Structured value: my-config: my-nested-field: field: "value"

It would seem reasonable IMO to basically remove structured and just change how we handle literals (basically replacing this block with case ResourceSelectorLiteral).

I thought about this option first but the reason why I've preferred to introduce the "structured" type was don't change the "literal" behavior, because it's already used by EnvRefs and SecretRefs fields, and for these use cases the "literal" option looks good enough (specially for env var case)

For these use cases just passing a literal, simple string value makes sense, but a structured, complex value does not, I think...so I didn't want to change that, and introduced the new "structured" option (I used this word because Pulumi docs also use it to reference about complex configs. So the main reason was backward compatibility.

but you're right, it sounds a bit strange. maybe we could handle "literal" for configs as you suggested and also keep the legacy "literal" behavior for the other Refs fields. WDYT?

Yes, we don't want to change literal handling for EnvRefs or SecretRefs. (There's an argument that we should, but we don't need to think about that as part of this feature.)

but you're right, it sounds a bit strange. maybe we could handle "literal" for configs as you suggested and also keep the legacy "literal" behavior for the other Refs fields. WDYT?

Yeah, that's what I was thinking. Let EnvRefs and SecretRefs continue to use resolveResourceRef, and you can use your structure handling for literals here.

ljtfreitas · 2024-05-14T16:43:49Z

thank you @blampe, don't worry, don't need to apologize, thank you for your support (i hope that I have not sounded like a jerk or something 😅 also), i will check the comments 👀

github-actions · 2024-05-17T13:10:52Z

PR is now waiting for a maintainer to run the acceptance tests. This PR will only perform build and linting.
Note for the maintainer: To run the acceptance tests, please comment /run-acceptance-tests on the PR

ljtfreitas · 2024-05-27T13:36:00Z

hi friends 👋 , any news here? cc @blampe

miguelslemos · 2024-05-30T00:04:24Z

Hi @blampe, this feature is crucial for us, could you take a look when you get time, please?

blampe

Sorry, have been OOO. I'm still a little uncertain about the type of ConfigRefs, and I worry we're duplicating some of SecretRefs. It would be nice to test merging behavior when keys collide.

Hi @blampe, this feature is crucial for us, could you take a look when you get time, please?

You should be able to use a custom image if this is blocking you.

blampe · 2024-05-29T14:12:43Z

pkg/controller/stack/stack_controller.go

+					allConfigs = append(allConfigs, configs...)
+				}
+			// Secret should be handled here as well because auto.ConfigValue should be marked as Secret:true
+			case shared.ConfigResourceSelectorType(shared.ResourceSelectorSecret):


When would I use this instead of the existing SecretRefs?

My original idea was to support the same options from ResourceRef but I'm wondering about that...maybe should we support just literal/structured values and ConfigMap for configs?

blampe · 2024-05-31T06:16:55Z

pkg/controller/stack/stack_config_test.go

+		sourceConfigMap := map[string]any{
+			"aws:assumeRole": map[string]any{
+				"roleArn":     "my-role-arn",
+				"sessionName": "my-session-name",
+			},
+			"aws:defaultTags": map[string]any{
+				"tags": map[string]any{
+					"my-tag": "tag-value",
+				},
+			},
+		}


For what it's worth, I think if you add another key like "aws:assumeRole.roleArn": "other-role-arn" this will fail non-deterministically.

Could be but, sorry, is it a real problem? For a stack configured like this,

aws:assumeRole: roleArn: "my-role-arn" aws:assumeRole.roleArn: "other-role-arn"

I wouldn't expect any deterministic behavior here, it would look like a configuration mistake to me.

I know it may seem weird, but it's possible so we need to consider it. Hyrum's law.

It would be very helpful to see some more tests around what happens when multiple config maps are provided with overlapping keys like this.

blampe · 2024-05-31T06:20:43Z

pkg/controller/stack/stack_controller.go

+	for _, configRefMap := range sess.stack.ConfigRefs {
+		for k, ref := range configRefMap {


I'm not clear on why this is a []map[string]ConfigRef instead of just a []ConfigRef. The key only seems to be used for secret refs (which mirrors SecretRefs?), and for the default case (do we need that?).

It was a design decision, that I've mentioned here: #575 (comment)

I've chosen the array of maps option to be similar to the envRefs and secretRefs fields, where the key is the env/secret key name.

But in fact for structured/config map cases the key is not used 😕...not sure what could be a better option. wdyt?

For literal values it's used as the config-key name:

configRefs: - sample: type: Literal value: just-a-value

it will be resolved as a config like that:

sample: just-a-value

blampe · 2024-05-31T06:23:02Z

pkg/controller/stack/stack_controller.go

+			default:
+				// try to resolve as a ResourceRef


Do we need this? resolveResourceRef doesn't support structure so it seems odd to default to it. An error would be reasonable IMO.

It doesn't, but resolveResourceRef supports all other cases like env, file, etc. These options are supported in configRefs too.

Maybe it does not make sense and configRefs should support just structured/config map options?

Yes, we should keep the new API surface area as small as possible. Unless we have a good reason to include them now, we can omit env, file etc. and add them later if needed. We don't need a key k for the literal case which leaves this default case as the only reason for it. Not needed IMO.

…nfigure Secrets through ConfigRefs

github-actions · 2024-06-03T23:28:26Z

PR is now waiting for a maintainer to run the acceptance tests. This PR will only perform build and linting.
Note for the maintainer: To run the acceptance tests, please comment /run-acceptance-tests on the PR

ljtfreitas · 2024-06-03T23:35:37Z

hey @blampe , I've made some changes and updated the PR description; I've removed the structured concept and started to handle just as literal values, both simple or complex ones.

blampe

I think we can still simplify the API a bit. Let's try to scope this down to only the things we know we need, which it sounds like is config maps and literals?

Other things that will help get this over the line:

Blocker: Resolve non-determinism by making ConfigRefs an array.
Blocker: Tests around edge cases with merging multiple refs/literals.
Optional: We could limit the change to an alpha API, to leave room for changes before we add it to v1.

blampe · 2024-06-21T17:28:48Z

deploy/crds/pulumi.com_stacks.yaml

+                  If this is omitted, configuration is assumed to be checked in and
+                  taken from the source repository. If present, ConfigRefs values


Does this mean we over-write in-repo configs entirely? Why not overlay these configs on top of source?

It seems surprising to be able to commit some changes to config, deploy it, but then not see the new config take effect. Is this the existing behavior of the operator?

blampe · 2024-06-21T17:32:31Z

deploy/crds/pulumi.com_stacks.yaml

+                          namespace:
+                            description: Namespace where the ConfigMap is stored.
+                              Deprecated; non-empty values will be considered invalid
+                              unless namespace isolation is disabled in the controller.
+                            type: string


Let's not add new things that are already deprecated.

Suggested change

namespace:

description: Namespace where the ConfigMap is stored.

Deprecated; non-empty values will be considered invalid

unless namespace isolation is disabled in the controller.

type: string

blampe · 2024-06-21T17:33:47Z

deploy/crds/pulumi.com_stacks.yaml

+                        description: Env selects an environment variable set on the
+                          operator process


Seems like this was copy-pasted, let's make sure the doc strings are accurate.

blampe · 2024-06-21T17:36:06Z

deploy/crds/pulumi.com_stacks.yaml

+                          of selector. Must be one of: Env, FS, Secret, ConfigMap,
+                          Structured, Literal'


Suggested change

of selector. Must be one of: Env, FS, Secret, ConfigMap,

Structured, Literal'

of selector. Must be one of: Env, FS, ConfigMap, or Literal.'

Or just ConfigMap/Literal if you scope it down to those two.

blampe · 2024-06-21T17:39:15Z

deploy/crds/pulumi.com_stacks.yaml

@@ -898,6 +975,83 @@ spec:
                  which can be optionally specified inline. If this is omitted, configuration
                  is assumed to be checked in and taken from the source repository.
                type: object
+              configRefs:


It was suggested that we keep this limited to an alpha version like this before adding it to v1. That way we aren't immediately locked in to the API.

blampe · 2024-06-21T17:52:55Z

pkg/controller/stack/stack_controller.go

@@ -1085,6 +1086,64 @@ func (sess *reconcileStackSession) SetEnvRefsForWorkspace(ctx context.Context, w
 	return nil
 }

+func (sess *reconcileStackSession) resolveConfigRefs(ctx context.Context) ([]ConfigKeyValue, error) {
+	allConfigs := make([]ConfigKeyValue, 0)
+	for _, configRefMap := range sess.stack.ConfigRefs {


This is still non-deterministic, and the name is still unused. Please make this an array.

blampe · 2024-06-21T18:10:21Z

pkg/controller/stack/stack_config.go

+	Value auto.ConfigValue
+}
+
+func NewStructuredConfigFromJSON(key string, rawValue apiextensionsv1.JSON) (*StructuredConfig, error) {


Don't we want to accept YAML?

blampe · 2024-06-21T18:13:12Z

pkg/controller/stack/stack_config_test.go

+		sourceConfigMap := map[string]any{
+			"aws:assumeRole": map[string]any{
+				"roleArn":     "my-role-arn",
+				"sessionName": "my-session-name",
+			},
+			"aws:defaultTags": map[string]any{
+				"tags": map[string]any{
+					"my-tag": "tag-value",
+				},
+			},
+		}


I know it may seem weird, but it's possible so we need to consider it. Hyrum's law.

It would be very helpful to see some more tests around what happens when multiple config maps are provided with overlapping keys like this.

blampe · 2024-06-21T18:16:09Z

pkg/controller/stack/stack_controller.go

+				literalRef := configRef.Literal
+				if literalRef != nil {
+					// ConfigLiteralRef handles both simple and structured values as json, flattening all keys to build a list of Pulumi key:value configs
+					structuredConfig, err := NewStructuredConfigFromJSON(k, literalRef.Value)


Already mentioned above, but interpreting literal as an object eliminates the need for k while still allowing structure.

literal: value: aws:region: us-west-2 foo:bar: - baz

blampe · 2024-06-21T18:21:52Z

pkg/controller/stack/stack_controller.go

+			default:
+				// try to resolve as a ResourceRef


Yes, we should keep the new API surface area as small as possible. Unless we have a good reason to include them now, we can omit env, file etc. and add them later if needed. We don't need a key k for the literal case which leaves this default case as the only reason for it. Not needed IMO.

ljtfreitas added 6 commits April 23, 2024 21:34

Initial implementation to support structured config

20eb3c7

Initial implementation to support ConfigRefs

d62e4b0

Reverting changes about Config; just ConfigRefs will be a complex config

d637e40

Trying to make tests pass...

8c8d103

Refactoring StructuredConfig + controller's support to ConfigRefs tests

79cadf2

Fixing some descriptions on CRD config fields

483bec5

ljtfreitas mentioned this pull request Apr 25, 2024

Support structured config #516

Closed

geowalrus4gh mentioned this pull request Apr 25, 2024

Support for structured config #258

Open

rquitales self-assigned this Apr 25, 2024

mikhailshilkov requested a review from blampe May 10, 2024 08:53

blampe reviewed May 10, 2024

View reviewed changes

rquitales assigned blampe May 15, 2024

Refactoring ConfigRefs to be an array instead of a map

98559a9

ljtfreitas requested a review from blampe May 20, 2024 19:59

mikhailshilkov mentioned this pull request May 28, 2024

Allow opt-out of adding stdout/stderr env variables for remote command pulumi/pulumi-command#451

Merged

blampe reviewed May 31, 2024

View reviewed changes

Renaming StructuredRef to ConfigLiteralRef and removing support to co…

b76e8b4

…nfigure Secrets through ConfigRefs

ljtfreitas requested a review from blampe June 3, 2024 23:36

blampe reviewed Jun 21, 2024

View reviewed changes

		for _, configRefMap := range sess.stack.ConfigRefs {
		for k, ref := range configRefMap {

		If this is omitted, configuration is assumed to be checked in and
		taken from the source repository. If present, ConfigRefs values

		description: Env selects an environment variable set on the
		operator process

		of selector. Must be one of: Env, FS, Secret, ConfigMap,
		Structured, Literal'

	of selector. Must be one of: Env, FS, Secret, ConfigMap,
	Structured, Literal'
	of selector. Must be one of: Env, FS, ConfigMap, or Literal.'

Support structured config #575

Are you sure you want to change the base?

Support structured config #575

Conversation

ljtfreitas commented Apr 25, 2024 • edited Loading

Context

Proposed changes

ConfigLiteralRef

ConfigMapRef

Configs merge strategy

Configs set strategy

Related issues (optional)

github-actions bot commented Apr 25, 2024

rquitales commented Apr 25, 2024

github-actions bot commented Apr 25, 2024

ljtfreitas commented Apr 25, 2024

blampe commented Apr 25, 2024

ljtfreitas commented Apr 26, 2024

blampe commented Apr 26, 2024

ljtfreitas commented Apr 27, 2024

ljtfreitas commented May 8, 2024

blampe commented May 9, 2024

blampe left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

ljtfreitas May 14, 2024 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

blampe May 15, 2024 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

ljtfreitas May 16, 2024 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

ljtfreitas May 14, 2024 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

ljtfreitas commented May 14, 2024 • edited Loading

github-actions bot commented May 17, 2024

ljtfreitas commented May 27, 2024

miguelslemos commented May 30, 2024

blampe left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

ljtfreitas Jun 3, 2024 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

github-actions bot commented Jun 3, 2024

ljtfreitas commented Jun 3, 2024

blampe left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

ljtfreitas commented Apr 25, 2024 •

edited

Loading

Configs `set` strategy

ljtfreitas May 14, 2024 •

edited

Loading

blampe May 15, 2024 •

edited

Loading

ljtfreitas May 16, 2024 •

edited

Loading

ljtfreitas May 14, 2024 •

edited

Loading

ljtfreitas commented May 14, 2024 •

edited

Loading

ljtfreitas Jun 3, 2024 •

edited

Loading