Currently, when there is a scenario stopped, global status of the campaign is KO.
If one scenario is stopped, we want now the global status of the campaign to be stopped in history.
SUCCESS < FAILED < STOPPED

Currently, if we rename a parameter, its a delete and create. So we have to rename manually in each parent or child component the inherited parameter.

@rbenyoussef
I let you fill this

Today we are able to produce a message to a kafka topic using the kafka-basic-publish task.

We need to add a new task to consume a message from a kafka topic in order to validate that we sent the intended data.

It would be nice to have some flexible inputs for the task as:

• a selector(json path like)
• commitSync(true|false)
• number of messages expected to return
• number of polls or timeout
• polling time

When we create a component, the page is not refreshed with the new id created.

Consequence :
We can not save modification on the component except if we reload the page.

On campaign page /#/campaign, make font-size smaller in order to deal with long campaign name.

Create an icon for campaign description like on scenario list.

On scenario execution history (.../#/scenario/16-13/execution/319), display execution environment

Ideas :

• Clickable component in scenario edition => DONE
• Replace the parameters with their values ​​in components and scenarios edition
• The task could directly propose their parameterization to avoid creating parent components
• It should be possible to overload the names of the components when they are used in parent components or scenarios
• Add Given/When/Then in component scenario edition

On scenarii page, show :

• the total number of scenarios
• the total number of filtered scenarios

On campaign reports, we want to display the number of success/fail scenarios

Today we are able to produce a message to a kafka topic using the kafka-basic-publish task.

We need to add a new task to consume a message from a kafka topic in order to validate that we sent the intended data.

It would be nice to have some flexible inputs for the task as:

• a selector(json path like)
• commitSync(true|false)
• number of messages expected to return
• number of polls or timeout
• polling time

We should add others status like stopped in campaign report label. Currently, there is only OK/KO

Backend : Create http controller:

• findAll(): returns dto with just: id, name, descripion, lastupdated dated
• findById(id): returns the entire dataset
• findVersion(vx): returns the entire datasetat version vx
• create(...)
• update(...)
• delete (by id)

Frontend : Create the service that will call the backend to retrieve the information from the dataset:

Persistence of datasetin a database (TBD)
History management

On scenario or campaign execution, add environment name in context and in global variable :
for context: #environment
for global var : environment

Currently, the button replay failed scenario does not appear when the campaign was stopped

Actually when combining tags eg : "tag1, tag2" it will create two tags "tag1" " tag2" with space at the beginning of tag2.

issue :

• step 1 has tags : tag1, tag2
• step 2 has tags : tag1,tag2

we will have 3 tags : "tag1", " tag2", "tag2"

Add user id who launch the execution of the scenario or the campaign

Thoughts about glacio integration problems.

The primary goal is to have a new syntax to write scenario which is well known in the dev world : cucumber feature file.

Architecture

Glacio could be viewed as an adapter to Chutney engine service, which is execution of a "step definition".
Therefore, this adapter should translate feature file into a step definition.

1. Glacio dependency
   We use for now the parser module of Glacio and depend only on the ast.
   Maybe ast model should be used instead to have fully support of Gherkin syntax as Background declaration for example...

2. Extension point
   One should be able to add the possibility to support new syntax.
   This could be done via implementing a spi interface which define two entries :

• Declaration of keywords to identify the service to use
• Transformation method to map a Glacio ast to Chutney engine step definition

3. Chutney dependencies
   Glacio adapter, being a separate module for now, must have a dependency on environment referential which is in server module.
   This seems wrong. Server module does too many things.
   This crying for new module extraction.

4. JUnit
   It would be nice if one could execute Glacio feature like cucumber does with its JUnit runner.
   Glacio has a junit runner module.
   Maybe we should have a Chutney runner.

Writing

Gherkin is the base syntax for Glacio extension. This latter extends it by adding the possibility to have multi-level step.

1. Step name
   Gherkin gives use the name for free
   To construct StepDefinition, one needs an environment.

2. Task
   Executable steps are leaf steps in the multi-level declaration of a scenario.
   We start defining special keywords in order to fully identify executable steps: Execute / Run.
   Those must start an executable step name.

3. Inputs / Outputs
   Task's Inputs/Outputs are identified through sub-steps with specific keyword: With.
   Gherkin DataTable/DocString could be used here to fulfill value.

4. Target
   Targets name are identified through sub-step with specific keyword: On.
   Only name has to be declared as targets should be provided resources.

5. Strategy
   ...

Create the Dataset menu

• add button to menu
• create the routing
• create an empty page

Create Dataset list page:

• edition button which routes to the Dataset editing screen
• creation button which routes to the Dataset editing screen (empty as it is a creation)
• click on a line displays a modal with a preview of the JDD (see wireframe)
• Search on name and desctiption
• Filter by tag combo box multi select as for campaign edition

Wireframe :

campaign run duration is the sum of all scenarii duration even in parallels run.

reproduction:

1. add 2 scenarii to campaign
2. set parallels exec mode
3. run campaign
4. compare the duration of the campaign execution with the real time taken

Example : scenario1 60s , scenario2 120s
expected : 120s
actual : 180s

To gain on adoption, it could be very useful for chutney to be able to invoke requests from a remote Postman collection.

I am not sure where it should be configured in Chutney, not sure about Target as a collection can be relative to several environments.
Say some API documentation (using swagger or equivalent) is hosted somewhere with its matching Postman collection;
Instead on the low-level HTTP task, a step can refer to a request in a collection that is evolving according to the documented Target's API.

A Postman collection is in fact a large Json file documenting how an API could be used with all elements needed to construct proper HTTP calls, see https://www.postman.com/collection for more details.

Postman collection may refer to variables, the same way Tasks already are.

To reproduce the bug :

• Go on execution history (.../#/scenario/16-13/execution/291)
• Click on "Edit"
• Click on "Save" or "Cancel"
• Close the browser tab
  => the pop up "Are you sure you want to close....." appears

Todo :
No pop up

To reproduce the bug :

• one component A
• one component B which have component A in its steps

On component A, in "used in step" on right panel, click on component B.
On left panel, click on component A
The click on component B in "used in step" on right panel does not work anymore.

When executing a scenario

• recover the associated dataset if it exists.
• Add the "pre-compilation" phase of the dataset just after that made for the parameters which will:
• replace unique values ​​like parameters
• If a step uses as a variable one of the header declared in the dataset , then the step is multiplied with for each iteration the replacement of the values ​​of each line of the dataset
• save the dataset used when executing the scenario

When running a campaign

• recover the associated dataset if it exists.
• at the execution of each scenario, pass the dataset which will overload the default dataset of the scenario
• save the dataset used during the campaign execution

Currently, in the retry strategy, we have a total time and a waiting time.
The waiting time corresponds to the waiting time after a failed step.
The total time will be used to define how many iterations there will be.
For example if

• total time = 20 min
• waiting time = 2 min,
  there will be a maximum of 10 iteration. No matter the execution time of the step. So in real time, we can exceed the total time.

The issue with this retry strategy is when the retry strategies are nested because the waiting times will be multiplied. For example if we have two nested strategies with max time = 10 min and waiting time = 1 min.
The maximum time will be 100 min.

When executing sql action from component screen, out of memory appears

${....}toto => failed
toto${...} => success

Should be ok in this two case

steps to reproduce :

1. access campaign page
2. edit campaign
3. cancel edition

expected : redirection to 'campaign/{ID}/execution'
current : redirection to '/campaign'

On administration -> target, target are sorted in firefox but no in chrome.

For all the screens :
if the scenario or campaign has no parameters, no need for JDD
Edit the controllers to add dataset reference persistence

Campaign execution report screen

• On a campaign report you can view the dataset used during execution (you can reuse the preview component made in the campaign list)

Campaign editing screen

• Add a combo box to select a default dataset for the campaign

Component type scenario editing screen

• Add a combo box which allows you to select a default JDD for the scenario

Scenario execution screen (component only)

• display what is the default JDD of the scenario
• add a combo box to select another dataset(which will not be saved) used if the user clicks on the execute scenario button

Campaign execution screen

• display what is the default datasetof the scenario
• add a combo box to select another dataset(which will not be saved) used if the user clicks on the button execute campaign

In campaign report screen
In scenario list screen

Create Dataset edition page

• reactiv form
• implement candeactivate if closing the page without saving
• create save and cancel button
• saving does not close the screen. A message appears saying that it has been saved successfully
• input name and description
• 4 tabs:
  • one for unique value
  • one for the datalist
  • one for links (not in wireframe)
  • one for history (not in wireframe)

Multi value tab:

• add line button which adds a line
• add header button which adds a header with an empty value for each line
• import / export according to the same rules as for the single value

Single value tab:

• add value button which adds a line
• on each line:
• input for the key
• input for value
• delete button
• Import which opens a selection of csv files. The import delete deletes / deletes / creates all the values ​​(to be confirmed, you can also just replace / add). The import is only done on the front side. It is up to the user to save next
• export create a csv file of unique values

Links
We can create a third "links" tab which will display where the Dataset is used (which scenarios and which campaigns). We could even display a link to the last 5/10 executions made with this Dataset .

History
We can create a fourth tab "history" which will display the history of modifications of this Dataset

Wireframe

We can see here https://github.com/chutney-testing/chutney/releases that several versions exist, however it is not clear why they exist.

What are the new features, bug or other changes that motivated the creation of a new release ?

This may be a good starting point for a newcomer to see a chronology and the main topics the team already worked on.

• The authorisations are stored in a file on the server authorization.json
• Create a page to edit the file in a menu Admin -> Authorization
• We use the json editor we already have to edit the file
• Only ADMIN user defined in configuration or in ldap can connect to the application without beeing known in authorization.json
• By configuration, we have to map the key of the admin role in the ldap and the ADMIN role in chutney. For example, in ldap, we could have CHUTNEY_ADMIN_USER which will be map to ADMIN profile in Chutney
• For first access, the username is insert in DEFAULT role
• We cannot save the file (with a message explaining why) :
  • If the role DEFAULT does not exist
  • If a user has more than 1 role
  • If an authorization does not exist
  • If the file is bad formatted

List of authorisation which can be affected to a role :

• Scenario
  • SCENARIO_READ
  • SCENARIO_WRITE
  • SCENARIO_EXECUTE
• Campaign
  • CAMPAIGN_READ
  • CAMPAIGN_WRITE
  • CAMPAIGN_EXECUTE
• Env/Target
  • ENVIRONMENT_ACCESS
• Global variable
  • GLOBAL_VAR_READ
  • GLOBAL_VAR_WRITE
• Dataset
  • DATASET_READ
  • DATASET_WRITE
• Components
  • COMPONENT_READ
  • COMPONENT_WRITE
• Admin (BDD, configuration, backup,...)
  • ADMIN_ACCESS

Example of authorization.json file :

{
  "roles": [
    { "name": "READ_ONLY",
      "rights": [
        "SCENARIO_READ",
        "CAMPAIGN_READ",
        "GLOBAL_VAR_READ"
        ]
    },
    { "name": "AUTOMATIONNER",
      "rights": [
        "SCENARIO_READ",
        "SCENARIO_WRITE",
        "SCENARIO_EXECUTE",
        "CAMPAIGN_READ",
        "CAMPAIGN_WRITE",
        "CAMPAIGN_EXECUTE"
        ]
    },
    { "name": "PRODUCT_OWNER",
      "rights": [
        "SCENARIO_EXECUTE",
        "CAMPAIGN_EXECUTE"
        ]
    },
    { "name": "DEFAULT",
      "rights": [
        "SCENARIO_READ"
        ]
    }
  ],
  "authorization" : [
    { "name": "READ_ONLY",
      "users": [
        "NB12345"
        ]
    },
    { "name": "AUTOMATIONNER",
      "users": [
        "MB12345",
        "AH12345"
        ]
    },
    { "name": "PRODUCT_OWNER",
      "users": [
        "RB12345"
        ]
    },
    { "name": "DEFAULT",
      "users": [
        "AA12345"
        ]
    }
    
    ]
}

‼️
We must first improve the readability of the component scenario so that a read-only user can properly read a scenario.

💡 for later :
Create a notification for the adminn of a new access
Allow to lock a specific component

We should keep the searches / filters when going back in the browser.
We should do that with uri.

On page ../#/campaign, make clickable the list of 10 last campaign execution.
When we click on an execution, we should go to the campaign execution report

We want an alert if we have modification before closing component scenario edit screen

can't use url in global var, when running test '\' is escaped twice.

Url in global var :

Url when running test:

abort a new retry if stopping of scenario execution is requested

Add user id who launch edit the component

Steps to reproduce bug

• Delete a step in a component or Scenario which has others steps with parameters declared after the deleted step
• Modify a parameter value
• Save

The parameter value is not updated upon refresh.

We could add graph on campaign report juste like jenkins one.

the environment remains empty when accesssing the scenario.

example from campaign or by url

May be add a tempo in UI to avoid the user to launch the scenario two times

Reproduction steps :

1. open component & set parameter default value
2. save component
3. run component
4. close execution result

Expected :

• parameter value set at step1

Current :

• parameter value are empty