Released: 2020-07-02
Please Note, this feature is only available when the new mapper component with Sailor
2.6.13version is installed in the system.
We introduce a new concept for error handling in the integration flows. Now you have a possibility to choose what to do with errors. In short you can handle those errors by building a workflow to catch them and store them in a most convenient way for you.
Before you would get error reports when you subscribe to receive them per email or if you would constantly checking your flows for errors on the platform UI by logging in.
Now you can do both of these if you still prefer it, but you can proactively handle those errors and decide what should happen with these errors and who should handle them.
To use and customise error handling choose the icon presented in the right bottom of the flow designer page after you configure the first step in the flow.

You can choose to configure it right away or wait when you finish the flow design and configure it to handle errors in one go since only one error handling is supported in this version.
Here is how the completed flow with error handling would look. Note, it handled the error in the third step.

In this example all errors in email components get redirected into a separate Google spreadsheet document.
In case when your custom component is not using a system-wide mapper component
right before to map the fields, then it will not show in the component chooser
as a method to process your errors. You can overcome this by upgrading your custom component Node.js Sailor to at least 2.6.13 version.
At this moment the components which do not require mapper component input to function would not show in the components list to choose.
At the moment the Custom Error Handling works for limited use cases and errors. For example it can not handle the following cases:
The incoming metadata from the Custom Error Handling is hard-coded. It has the following structure:
{
"error": {
"message": "Error message",
"name": "Error name",
"stack": "Error stack trace"
},
"errorInput": {
"properties": {
"headers": { "header":"value" }
},
"content": {
"headers": { "header":"value" },
"body": {}
}
}
}
We continue our improvements and developments of the New Secrets Management service. In this round we made great improvements on authentication parameters.
auth_uri and token_uri for OAuth2 (v20.27)Authentication clients for OAuth2 now support auth_uri and token_uri attributes
in credentials.
Please Note: API-requests for auth-clients is different now. More iformation in auth-clients-(experimental) section of API documentation.
Secrets Management Service refreshes tokens 60 second earlier than
exires_in parameter record to ensure the stability of the connection.
expires_in taken from token_expires_in (v20.27)Now when you create a new secret linked to auth-client with a parameter token_expires_in,
the parameter expires_in gets the value from the token_expires_in data.
From 1.1.0 version of Request-Reply component you
can add and configure the HTTP response code using the new Response Code
input field during the design phase. Response code is passed in the reply message
header. The default value is 200 by default.
2.6.13.2.6.13.2.6.13.2.6.10.2.6.9.Released: 2020-07-30
We are introducing a new header for the flow designer page to maximise the screen real estate and separate the integration flow graph from the functional buttons.
The animation below shows all the functionality in one go.
We are introducing a new functionality to copy flows from one workspace to the another. You must have at least Integrator level access in both workspaces for this to work. The new copy flow functionality behaves the following way:
In our back-end we have introduced API endpoints to use with copy flow functionality. These endpoints are still in active development and should be regarded as experimental. Check our API-docs documentation pages for more details.
Introducing a new VPN Agent. It is using а VPN tunnel specifically configured to access any local resource in your local environment. It could be any local database or CRM to which you would rather restrict access from outside world but still need to query for a data. Agents are available to configure by visiting the Agents page as usual. The dedicated VPN Agents page has more details about this.
Starting from this release the old Local Agents will be deprecated. You can
still visit and configure your agents by removing the /vpn part of the address
while visiting the VPN Agents page.
We are introducing a new functionality called Lightweight messages. The main idea of this functionality is to reduce the load of messaging queue by sending the Lightweight messages through it. By reducing the load on messaging queue we reduce the processing time significantly and reduce or prevent out of memory errors during the data exchange between the steps.
Our aim for this realisation was to make sure the large size messages and attachments bypass the messaging queue all together. Instead of passing large messages through messaging queue we introduced an internal service which would temporarily store large message and provide an ID of these messages instead. These IDs, in turn, get sent through the messaging queue - not the big chunks of data.
If the step requires stored data it communicates with the storage directly, not through the messaging queue.
The Lightweight messaging is already implemented at system level. If you would like to use this functionality you must update the sailor version in your components to Node.js Sailor to
2.6.14version and Java Sailor to at least3.2.0version.
We continue our improvements and developments of the New Secrets Management
service. In this round we made great improvements authentication client (auth-client)
creation process. Previously auth-client creation was possible on tenant level
and only by system administration. In this release auth-client creation was extended
following way:
auth-client at workspace, contract and tenant level. This enables scopes according to the auth-client creation and dependence. For example the tenant auth-client is visible and can be used in secret in all workspaces in this tenant and so on.GET /v2/auth-clients is introduced to query values. You will get results only when you have permission to view these auth-clients. For example if you have only WORKSPACE.AUTH_CLIENT.GET permission you will only get workspace level auth-client records.GET /v2/auth-clients/:id is introduced to query auth-client endpoint with specific IDs. The standard access permissions apply in this case as well.auth-client record on workspace, contract and tenant level.Check our API-docs documentation pages) for more details.
2.6.14 and Java Sailor 3.2.0 versions which support the Lightweight messages functionality.3.3.0 version which supports the Custom Error Handler functionality.In our second iteration of Custom Error Handler feature we have made improvements and added more functionality.
Component container has failed to start errors as well.3.3.0 version.Appbuilder problem by using the current latest LTS Ubuntu version in the dockerfile.c/ in browser window.The component has undergone major refactoring.
2.6.14.this.logger instead of console.log.2.6.14.2.6.13.Released: 2020-08-27
With this release we have revitalised the Recipes functionality of our platform. It underwent changes under the hood (the Recipes API) as well as in the UI. For now the feature is in rapid advancement/development phase. Not all API endpoints have their equivalent UI solution on this stage.
You can export any flow to recipe from the UI:

When recipe is created you are redirected to the Recipe Activation page from where you can activate and use your recipe in any workspace where you have access provided that your newly created recipe has matching visibility level.
Check our Creating and Using Recipes page for more information.
After initial feedback from many customers we re-introduced back the original Copy Flow functionality. It will create an exact copy of your flow in the same workspace. You are presented with a pop-up to enter a new name for your copy.
We introduce the Export Flow functionality to create copy of your flow to different workspace where you have access. This replaces the previously introduced Copy flow functionality. Please visit our Copy and Export flow documentation page for more information.
The Custom Error Management icon is changed to info icon to better reflect the functionality.
Introducing a new component to work with Google BigQuery.
Query action to perform user-provided query.Split on JSONata Expression action with a possibility to emit non-array objects.2.6.14.OrderBy is not supported in One Drive for Business.2.6.14.Bulk Create/Update/Delete action with Bulk Upsert feature.Bulk action causing 404 - File's metadata is not found error.application/x-www-form-urlencoded encoding bug.Released: 2020-09-24
With this release we enabled lowering the recipe visibility using an API call.
To lower the visibility use the /v2/recipes/{RECIPE_ID}/visibility API endpoint.
Check our API documentation for exact API call format.
To remind, the visibility can be workspace, contract and tenant. The workapace
is the lowest visibility. Please note the following before trying to lower the visibility:
We extended the recipe creation API to include the author attribute as well.
Here is an example section of the recipe draft:
{
"data": {
"type": "recipe",
"attributes": {
"info": {
"title": "My simple recipe",
"author":"THE AUTHOR OF RECIPE NAME IS LOCATED HERE",
"short_description": "The short description text is here",
"description": "Long description text goes here."
},
With this release we improved the conditions when the polling flows start after you update and publish the draft which replaces the active flow.
When you publish a new draft over the active flow it would not trigger a new execution, but will wait and run it on next due execution time.
Please note: If you need to execute your flow right away you can always press Run now to do so.
We started an audit of code in our components against the possibility of them exposing sensitive data in the logs. The following components were checked and potential parts refactored:
Introducing new component Oracle JDE Edwards EnterpriseOne Orchestrator.
Execute an Orchestration actionQuery action metadata with additional options