feat:delete environment command

[shivam-Purohit]

Currently there is no command to delete the chaos environment. This feature will aim to add command to delete chaos environments based on the id provided.

[SarthakJain26]

@Shivam you will have to add a check to see if there are any infras that are present for the environment, if yes you can just prompt that there are infras present in the environment and they need to be deleted before deleting the environment.

[shivam-Purohit]

@SarthakJain26 @neelanjan00
Currently i am an issue. If i delete a environment and try to create a new environment immediately then this error occurs
Chaos Environment connection failed: write exception: write errors: [E11000 duplicate key error collection: litmus.environment index: environment_id_1 dup key: { environment_id: "env1" }]
does this is not reflecting in the db?

also we are using LIST environment. But as i checked there already exist a handlers for get environment in graphlql. I suggest we should use that for get chaos-environment command rather than iterating through list of environments.This way we would provide more details and isolate both commands, this will also help in testing commands separately. We should make changes in the get chaos-environment command andlist chaos-environmentscommand

	var isEnvExist = false
	for i := range Env.Data.ListEnvironmentDetails.Environments {
		if newInfra.EnvironmentID == Env.Data.ListEnvironmentDetails.Environments[i].EnvironmentID {
			utils.White_B.Println(Env.Data.ListEnvironmentDetails.Environments[i].EnvironmentID)
			isEnvExist = true
			break
		}
	}

This is other instance we are using the iteration which imo should not be needed.

[SarthakJain26]

@SarthakJain26 @neelanjan00 Currently i am an issue. If i delete a environment and try to create a new environment immediately then this error occurs Chaos Environment connection failed: write exception: write errors: [E11000 duplicate key error collection: litmus.environment index: environment_id_1 dup key: { environment_id: "env1" }] does this is not reflecting in the db?

also we are using LIST environment. But as i checked there already exist a handlers for get environment in graphlql. I suggest we should use that for get chaos-environment command rather than iterating through list of environments.This way we would provide more details and isolate both commands, this will also help in testing commands separately. We should make changes in the get chaos-environment command andlist chaos-environmentscommand

	var isEnvExist = false
	for i := range Env.Data.ListEnvironmentDetails.Environments {
		if newInfra.EnvironmentID == Env.Data.ListEnvironmentDetails.Environments[i].EnvironmentID {
			utils.White_B.Println(Env.Data.ListEnvironmentDetails.Environments[i].EnvironmentID)
			isEnvExist = true
			break
		}
	}

This is other instance we are using the iteration which imo should not be needed.

@shivam-Purohit On deleting it does get reflected in the DB but the error you are seeing is because we are not deleting the document from the DB rather we are just marking it as removed, and since an mongo index is set on name field, therefore mongo does not allows the creation of a new environment with the same name. The solution to this is make the index as partial index and not consider the documents that are marked as removed. Anyways this is a separate issue which will have to be updated in graphql code.

Regarding using GET command instead of LIST, you are right, we should use GET where we want to fetch just one resource and we should be using LIST where multiple resources are being fetched. You can go ahead and update the implementation. Thanks

[shivam-Purohit]

@SarthakJain26 so what should be the next step in this feature then?

Also I have pushed the changes I suggested if you can take a look.

[SarthakJain26]

@shivam-Purohit please also add the example usage of command in delete/delete.go and Usage_0.23.0.md file

[shivam-Purohit]

@shivam-Purohit please also add the example usage of command in delete/delete.go and Usage_0.23.0.md file

done! can you take another look? thank you!

[shivam-Purohit]

The solution to this is make the index as partial index and not consider the documents that are marked as removed.

I see a commit pushed in litmus regarding this. Should this issue now be fixed @SarthakJain26 ?

[SarthakJain26]

The solution to this is make the index as partial index and not consider the documents that are marked as removed.

I see a commit pushed in litmus regarding this. Should this issue now be fixed @SarthakJain26 ?

Yes Shivam! This will be fixed now.

[shivam-Purohit]

@SarthakJain26 done!