feat: Multi agent #5844

exu · 2024-09-11T10:18:39Z

Pull request description

Checklist (choose whats happened)

Breaking changes

Changes

Fixes

exu · 2024-09-26T13:54:27Z

cmd/api-server/main.go


 	// Check Pro/Enterprise subscription
 	var subscriptionChecker checktcl.SubscriptionChecker
 	if mode == common.ModeAgent {
 		subscriptionChecker, err = checktcl.NewSubscriptionChecker(ctx, proContext, grpcClient, grpcConn)
 		exitOnError("Failed creating subscription checker", err)
+
+		// Load environment/org details based on token grpc call
+		environment, err := controlplanetcl.GetEnvironment(ctx, proContext, grpcClient, grpcConn)


executor need to follow loading it from this one - not from inlined env variables

exu · 2024-09-26T13:55:05Z

pkg/agent/agent.go

 func (ag *Agent) executeCommand(ctx context.Context, cmd *cloud.ExecuteRequest) *cloud.ExecuteResponse {
 	switch {
-	case cmd.Url == healthcheckCommand:
+	case cmd.Url == HealthcheckCommand || cmd.Command == string(cloud.HealthcheckCommand):


This is not needed anymore

exu · 2024-09-26T13:56:42Z

pkg/agent/handlers/response.go

@@ -0,0 +1,47 @@
+package handlers


would it be better to have some errors wrapped
e.g.

errors.Wrap(err, errors.NotFound)

and next the mapper which will check what kinds of error this is and set valid response

exu · 2024-09-26T13:58:13Z

pkg/testworkflows/testworkflowexecutor/executor.go

@@ -542,11 +532,55 @@ func (e *executor) Execute(ctx context.Context, workflow testworkflowsv1.TestWor
 		log.DefaultLogger.Errorw("failed to encode tags", "id", id, "error", err)
 	}

+	// Get (for centralized mode) TW execution or create it
+	if request.Id != "" {


We need to de-concrete executions logic from here to split fully executor engine from data implementation

exu · 2024-09-26T14:00:21Z

pkg/testworkflows/testworkflowexecutor/executor.go

+	execution.Tags = tags
+
+	// Insert or save execution
+	if request.Id != "" {


exctract it outside?

executor could be in form

func(ExecutionRequest, Workflow) Result

rangoo94 · 2024-09-26T14:05:45Z

pkg/agent/handlers/execute.go

+
+// TODO - valid error handling
+
+func NewExecuteTestWorkflowHandler(


This is for sure not needed:

There is already option to execute Test Workflow via generic gRPC command we have (that calls API)

For future Execution Worker, the command will look differently - as the Execution Worker needs to be stateless

I've planned here to fully decouple from API server - as it's additional unneded thing. So it's quite needed to not spawn this APIServer

These was the attempt to decouple it from the API server totally - can be refactored/reordered if we decide about API later

I've planned here to fully decouple from API server - as it's additional unneded thing. So it's quite needed to not spawn this APIServer

But when it will be decoupled from the API Server, the signature will also differ, so (A) this function is for decoupling, yet (B) it needs to be deleted (and replaced with a new handler) after decoupling, as it will look differently 🙂 So probably better to not pollute the new gRPC schema with obsolete functions

rangoo94 · 2024-09-26T14:08:37Z

Considering that in this PR the runner IDs are not dynamic (but needs to be pre-created), either:

There should be no Runner ID thing, but rather separate Agent Keys for each of them
The Runner ID should be completely dynamic (and informal + not unique)

On the other hand, the best solution is to avoid runner IDs at all, and have runner tags instead (like K8S nodeAffinity -> Testkube runnerAffinity). It would work great:

simple,
Kubernetes-like,
fully customizable,
scalable (as runners could be created on demand too)
runnerAffinity could be defined on any level (globally, per workflow, per execution)

exu · 2024-09-27T06:49:27Z

Considering that in this PR the runner IDs are not dynamic (but needs to be pre-created), either:

There should be no Runner ID thing, but rather separate Agent Keys for each of them

The Runner ID should be completely dynamic (and informal + not unique)

On the other hand, the best solution is to avoid runner IDs at all, and have runner tags instead (like K8S nodeAffinity -> Testkube runnerAffinity). It would work great:

simple,

Kubernetes-like,

fully customizable,

scalable (as runners could be created on demand too)

runnerAffinity could be defined on any level (globally, per workflow, per execution)

But you need some kind of ID here (whatever we name it) to e.g. schedule against it - I'm not sure if we really should follow Kubernetes naming here at all. These points are both valid for IDs and tags like in affinity.

I agree about separate keys - if we want to decouple runners from environments - it'll be next thing to do for sure.

rangoo94 · 2024-09-27T13:22:12Z

But you need some kind of ID here (whatever we name it) to e.g. schedule against it

You only need to target a tag, not ID - if someone wants to have actually unique way to identify them, simply there may be some tag unique 🙂
IDs may be required, but not in the way that User needs to pass them and register in the Cloud. They may only be useful to pull the already started execution information, or control it (pause, resume, abort), It would be enough to provide such ID on the Agent (or rather Worker/Runner) only then (and associate to i.e. execution)

exu force-pushed the sandbox/multiagent branch 3 times, most recently from 0d0d34c to fc3c3af Compare September 18, 2024 08:10

exu changed the base branch from develop to main September 20, 2024 05:53

feat: added runner handlers to the agent

6e13778

exu force-pushed the sandbox/multiagent branch from 4c4d7a7 to 6e13778 Compare September 26, 2024 06:00

exu added 3 commits September 26, 2024 08:05

fix: golang ci fixes

8cf8646

fix: openapi spec fixes

dd4715b

fix: cleaning unused code

b38447d

exu marked this pull request as ready for review September 26, 2024 13:53

exu requested a review from a team as a code owner September 26, 2024 13:53

exu commented Sep 26, 2024

View reviewed changes

rangoo94 reviewed Sep 26, 2024

View reviewed changes

exu closed this Nov 28, 2024

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

feat: Multi agent #5844

feat: Multi agent #5844

exu commented Sep 11, 2024

exu Sep 26, 2024

exu Sep 26, 2024

exu Sep 26, 2024

exu Sep 26, 2024

exu Sep 26, 2024

rangoo94 Sep 26, 2024

exu Sep 27, 2024 •

edited

Loading

exu Sep 27, 2024

rangoo94 Sep 27, 2024 •

edited

Loading

rangoo94 commented Sep 26, 2024 •

edited

Loading

exu commented Sep 27, 2024

rangoo94 commented Sep 27, 2024 •

edited

Loading


		// TODO - valid error handling

		func NewExecuteTestWorkflowHandler(

feat: Multi agent #5844

feat: Multi agent #5844

Conversation

exu commented Sep 11, 2024

Pull request description

Checklist (choose whats happened)

Breaking changes

Changes

Fixes

exu Sep 26, 2024

Choose a reason for hiding this comment

exu Sep 26, 2024

Choose a reason for hiding this comment

exu Sep 26, 2024

Choose a reason for hiding this comment

exu Sep 26, 2024

Choose a reason for hiding this comment

exu Sep 26, 2024

Choose a reason for hiding this comment

rangoo94 Sep 26, 2024

Choose a reason for hiding this comment

exu Sep 27, 2024 • edited Loading

Choose a reason for hiding this comment

exu Sep 27, 2024

Choose a reason for hiding this comment

rangoo94 Sep 27, 2024 • edited Loading

Choose a reason for hiding this comment

rangoo94 commented Sep 26, 2024 • edited Loading

exu commented Sep 27, 2024

rangoo94 commented Sep 27, 2024 • edited Loading

exu Sep 27, 2024 •

edited

Loading

rangoo94 Sep 27, 2024 •

edited

Loading

rangoo94 commented Sep 26, 2024 •

edited

Loading

rangoo94 commented Sep 27, 2024 •

edited

Loading