Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

...

Option #1

Description

The hydrator app needs to be able to write/read to a dataset to store and retrieve drafts and other information about business logic.  We can implement a Hydrator CDAP Application with a service that can have REST endpoints to serve the required hydrator functionalities. Enabling Hydrator in a namespace will deploy this Hydrator app and start the service. Hydrator UI would ping for this service to be available before coming up. The back-end business logic actions which directly needs to use the CDAP services endpoints can be made generic. 

  • Pros

    • Everything (Drafts, etc) stored in the same namespace, proper cleanup when namespace is deleted.
  • Cons

    • Every namespace will have an extra app for supporting hydrator if hydrator is enabled. Running this service, will run 2 containers per namespace. we can add an option to enable/disable hydrator if we are not using hydrator in a namespace.  It might feel weird as a user app, as the user didn't write/create this app.  

 


Option #2

Description

We will still use an Hydrator CDAP app but we create an "Extensions" namespace and have the "hydrator" app only deployed in the "extensions" namespace, this app would serve the hydrator requests for all namespaces.

It will use a single dataset to store the drafts, row keys can be name spaced for storing the drafts, while deleting the namespace, the rows belonging to the namespace will be deleted from the dataset.
  • Pros

    • Less amount of resources used, only 2 container's used rather than 2 container’s per namespace, only one dataset is used.
    • Only one app for using hydrator across namespace and not an app per namespace, less clutter. 
    • New extensions could be added to the same namespace to support other use cases in future.
  • Cons

    • Using a single dataset for storing all drafts across namespace is less secure?.
    • User won't be able to create a new namespace called "Extensions", as it will be reserved.

Open Questions

  • How to delete the drafts when the namespace is deleted ?
  • When to stop this service? 
  • Availability of the service? 
  • Security
    • If we decide to add more capability in hydrator back-end app, Eg: Make the pipeline validation/deploy app, etc,  then in secure environment, 
    • The hydrator-service can discover appropriate cdap.service and call appropriate endpoints?

Option #3 (based on discussion with terence)

1) No new user level apps are deployed. Config store is used to store user drafts of hydrator apps.

2) REST endpoint 'configure', can accept partial config and return a config response with suggestions of values for fields in a plugin, exceptions if any during configuring the plugin. 

  • user can choose a value from the suggestions for the field and call the configure again.
  • user can look at exception, fix the issue with either the script or configuration and call configure again. 
  • when all the required configs are provided and there aren't any exceptions, completionStatus would be set to true for the plugin.

    #3 

    Story 1 - Schema and field value suggestions : 

    Plugin annotation @Endpoint: 

    Plugin’s can have custom plugin-specific methods that can be annotated as @Endpoint.
    UI can learn about available endpoints for a plugin from the plugin properties.
    UI can call the app-fabric endpoint identifying {artifact, plugin} with the method name and method-parameter as request body,  the app-fabric endpoint will then load the corresponding plugin and call the method identified by method-name, if the method is annotated as @Endpoint.
    The response from this method call is sent as response of the HTTP request.

    REST API :

    "/namespaces/{namespace-id}/artifacts/{artifact-name}/versions/{artifact-version}/extensions/{plugin-type}/plugins/{plugin-name}/versions/{plugin- version}/methods/{plugin-method}?scope={artifact-scope}&pluginScope={plugin-scope}"

    BodyMap<String, String> –  fieldName to value mapping.

    Example Methods in Plugin DBSource:

    @Endpoint
    List<String> listTables(Map<String, String> inputConfig)
    @Endpoint
    Map<String, String> getSchema(Map<String, String> inputConfig)


    Schema Propagation and Validation through backend - DryRuns:

    Currently when pipeline is published, configurePipeline of plugins are called and we perform pipeline validation and plugin validations and also deploy the application. 
    1.Goal of dry-run endpoint is to validate a pipeline, then validate plugins by calling configure methods of plugin’s in the pipeline without
       performing any creation of datasets or generate program etc, which are usually done during deploy.
    2. using dry-run we would be able to catch issues in the pipeline earlier and fix them before deploying. 
    Dry-run can also be used by UI for schema propagation with some requirements from UI:
    •      If Plugin has field “schema", UI can mutate the output schema
    •      If plugin doesn’t have the field “schema" , UI cannot change the output schema and has to rely on result of dry               
                   run for the output schema for that stage, which is set during plugin configuration.
    we need to follow the above conditions for correctness, if UI mutates schema when there isn’t a field “schema”, the backend would have a different schema as input-schema for the next stage and the UI changes wouldn’t be reflected.
    REST : namespace/{namespace-id}/dry-run 
    Request-Body : JSON Config.
    Response-Body: JSON Config with additional fields in the plugin for output schema, exceptions in configuring pipeline stage, etc.  


    Story 2 - Drafts

     

    HTTP Request Type

    Endpoint

    Request Body

    Response Status

    Response Body

    POST

    /namespaces/{namespace-id}/configurations/{config-id}/


     

    {

    "config": {...}

    }

    200 OK: config saved successfully

    409 CONFLICT: draft-name already exists

    500 Error: while saving the draft

     

    PUT

    /namespaces/{namespace-id}/configurations/{config-id}/


     

    {

    "config ": {...}

    }

    200 OK: config updated successfully

    404 NOT Found : config doesn't exist already, cannot be updated.

    500 Error while updating the config

     

    GET

    /namespaces/{namespace-id}/configurations/{config-id}/

     

    200 return all the versions for the config identified by the config-name

    404 config not found

    500 error while getting config


     

    [

    {

    "timestamp" : "...",

    "config": {

    "source" : {

       ....

     },

    "transforms" : [...],

    "sinks" [...]

    "connections" : [..]

    }

    },

    ...

    ]

    GET

    /namespaces/{namespace-id}/configurations/{config-id}/versions/{version-number}

    -1 -> latest version

     

    200 return the versions for the config identified by the config-id and version-number

    404 config with version found

    500 error while getting config


     

    {

    "timestamp" : "...",

    "config": {

    "source" : {

       ....

     },

    "transforms" : [...],

    "sinks" [...]

    "connections" : [..]

    }

    }

    GET

    /namespaces/{namespace-id}/configurations/

     

    200 return the name of list of all saved configs

    500 error

    [
     "streamToTPFS",
     "DBToHBase",
      ...
    ]

    DELETE

    /namespaces/{namespace-id}/configurations/

     

    200 successfully deleted all configs

    500 error while deleting

     

    DELETE

    /namespaces/{namespace-id}/configurations/{config-id}

     

    200 successfully deleted the specified config

    404 config does not exist

    500 error while deleting

     

     

    The ConsoleSettingsHttpHandler currently makes use of ConfigStore. It's however not name-spaced and has few other issues, it can be fixed and can be improved to store configs.

    Along with pipeline drafts ConsoleSettingsHttpHandler also stores the following information currently:

    Code Block
    titlePlugin Template Endpoints
    GET namespaces/{namespace-id}/plugin-templates/{plugin-template-id}/ 
    // create a new plugin template
    POST namespaces/{namespace-id}/plugin-templates/{plugin-template-id}/ -d '@plugin-template.json' 
    // update existing plugin template
    PUT namespaces/{namespace-id}/plugin-templates/{plugin-template-id}/ -d '@plugin-template.json'
    // delete the plugin template
    DELETE namespaces/{namespace-id}/plugin-templates/{plugin-template-id}/ 
    Code Block
    titleDefaults
     // create/update defaults this include user's plugin version preferences, etc.
     PUT : namespaces/{namespace-id}/defaults -d '@default.json' 
     GET : namespaces/{namespace-id}/defaults 

     

    JAVA API - Config Store:

    Code Block
    titleExisting configstore methods
    void create(String namespace, String type, Config config) throws ConfigExistsException;
    
    void createOrUpdate(String namespace, String type, Config config);
    
    void delete(String namespace, String type, String id) throws ConfigNotFoundException;
    
    List<Config> list(String namespace, String type);
    
    Config get(String namespace, String type, String id) throws ConfigNotFoundException; 
    
    void update(String namespace, String type, Config config) throws ConfigNotFoundException;
    Code Block
    titleConfigstore new methods
    // get a particular version of an entry. 
    Config get(String namespace, String type, String id, int version) throws ConfigNotFoundException; 
    // get all the versions of an entry.
    Config getAllVersions(String namespace, String type, String id) throws ConfigNotFoundException; 
    // delete all entries of specified type.
    void delete(String namespace, String type) 

    Open Questions :

    1) ConfigStore stores the configs in "config.store.table", currently the table properties doesn't have versioning, drafts would need versioning, this would also need CDAP-upgrade to update properties for the existing dataset? 

    2) rename ConsoleSettingsHttpHandler to ConfigurationsHttpHanlder ?

    3) Dependent UI changes. 

    User Stories (3.5.0)

    1. For the hydrator use case, the backend app should be able to support hydrator related functionalities listed below:
    2. query for plugins available for a certain artifacts and list them in UI
    3. obtaining output schema of plugins provided the input configuration information
    4. deploying pipeline and start/stop the pipeline
    5. query the status of a pipeline run and current status of execution if there are multiple stages.
    6. get the next schedule of run, ability to query metrics and logs for the pipeline runs.
    7. creating and saving pipeline drafts
    8. get the input/output streams/datasets of the pipeline run and list them in UI. 
    9. explore the data of streams/datasets used in the pipeline if they are explorable. 
    10. Add new metadata about a pipeline and retrieve metadata by pipeline run,etc.
    11. delete hydrator pipeline
    12. the backend app's functionalities should be limited to hydrator and it shouldn't be like a proxy for CDAP.  

    Having this abilities will remove the logic in CDAP-UI to make appropriate CDAP REST calls, this encapsulation will simplify UI's interaction with the back-end and also help in debugging potential issues faster. In future, we could have more apps similar to hydrator app so our back-end app should define and implement generic cases that can be used across these apps and it should also allow extensibility to support adding new features. 

    Generic Endpoints

    ...