Fixed Retrieve Duplicate Database IStructure with Primary Key Data Types #
There is an issue when generating the IStructure from the MySQL database. A duplicate column IStructure with primary key type is being generated from the New Agent, leading to the transfer of the duplicate generated IStructure to Salesforce via the Agent Database Adapter.
Fixed Not Unchecked Initial Mode After do Outbound Processing with the New Agent #
An issue was encountered during outbound processing in the initial mode with the new agent. Upon completion of the outbound processing, the Initial Mode flag persists. This problem has been resolved by unchecking the Initial Mode flag after processing is complete.
Agent File Continues Processing While Incorrectly Setting the Path Backup Folder of the Agent File Adapter #
This issue occurred when using a processing file with the Agent File Adapter of the New Agent and incorrectly setting the property adapter for the Backup Folder. In this scenario, the agent would not process properly and enter a loop, repeatedly attempting the task. We resolved this by automatically creating a backup of the folder specified in the Backup Folder input. In case of processing failure, the file will be moved to the error Folder created by the New Agent.
Fixed: Cannot Delete Child Records from Database with Hierarchical V3 Outbound Interface Design #
There is an issue when performing hierarchical V3 outbound processing of the Agent Database Adapter with a delete operation to remove records from the database. The second sub-interface level processes incorrectly, as it derives the table name from the parent interface, resulting in erroneous deletion of child records. This problem has been resolved by accurately specifying the table name to formulate the delete query.
Fixed Exporting the XML File Containing the Interface Hierarchy Designed with Interface v2 in the old Style #
The issue arises when conducting outbound processing to export an XML file with Agent File Adapter containing the interface hierarchy designed with interface v2 in the old style. The XML file generated an incorrect hierarchical data payload. This issue has now been resolved.
Fixed Job Status ‘FAILED’ Issue When Importing CSV Files in Skyvva Bulk Mode #
There is an issue when processing in Bulk Mode with the new Agent with CSV File Type. Although the Basket has already been created, the Job Status shows ‘FAILED’ with the Job Comment stating ‘Process is Failed’ for both manual and scheduler processing. This problem has been resolved by updating the correct Job status and Job Comment to ‘Process is Completed’ upon completion.
Resolved Bug: Agent Connection Status with Case Sensitivity Ignored #
An issue arises when a user creates an Agent Connection with a non-open firewall on the Salesforce side, resulting in an inactive status with a lowercase. When pinging the connection and attempting to resubscribe the Agent Connection with the “Non-open Firewall” button enabled in the “Register Agent Connection” section of the Agent Web App, an unexpected error may occur. To address this, we will resolve the issue by ignoring the case sensitivity.
Encountering Duplicate Root Messages When Using the Custom Processing Class Payload in the Agent File Adapter #
There is an issue when customizing the payload within the Agent File Adapter of the Custom Class Payload. Upon processing, you will receive two root messages: one being a custom root message and the other being the Agent Message.
Resolved Issue with a Request and Getting Two Response Records #
The issue stemmed from making a record request that returned two records with the same request. Nevertheless, the child records failed to generate a correct tree structure for each parent record. This problem has been resolved to ensure that all child records are processed correctly with their respective parent records.
Resolved issue with Agent Database Adapter Processing using Transactional Handling #
An issue occurred during the processing of the Delete Operation using the Agent Database Adapter with Transactional Handling Mode in Interface V3 Design. Despite this, the record was not deleted, resulting in a failed message with the error exception ‘Read timeout, Line: 2314, Trace: C…’ in synchronous mode, and a pending message in asynchronous mode processing. This issue has been resolved, now allowing successful deletion and updating of messages upon record deletion.
Fixed Issue with Custom Format with Agent File Adapter with Non-Open Firewall #
There was an issue during outbound processing with the Agent File Adapter (Content Format = Custom and Custom Processing Class) in non-open firewall mode. It faced difficulties creating a custom message without a business message and an agent message. This problem has been resolved to prevent the duplication of root messages and ensure the creation of only a Custom Message to cover the business message.
Fixed Agent Rest Adapter Generate incorrect Payload #
There are a few issues regarding generating the payload through the Agent Rest Adapter as follows:
- Encounter an issue while mapping an sObject to the request body of the REST message template using the Agent Rest Adapter. After the Agent processing, an error message is logged in the agent log: ‘java.lang.Exception: Root message is not allowed to map with the “Body” node. We need to map the business field or node below the “Body” node.’. This problem has been rectified to allow direct mapping with the request body for processing purposes, enabling the generation of the payload as usual.
- The issue encountered during mapping the business node under the request body of the REST message template has been resolved. After Agent processing with the Agent Rest Adapter, a duplicate business node was found in the JSON payload. This problem has been rectified by ensuring the generation of only one business node, as specified in the defined mapping.
Fixed Agent Handling of Skyvva App Version Values During Processing #
We have improved the New Agent to manage the Skyvva App Version when the correct version is entered without a dot. This issue could occur during processing and lead to the error ‘For input string: 251.9’ in the post-processing log. The problem has been resolved as outlined below:
- When the input is “251.9”, the conversion result is “2519”, enabling the Agent to recognize this as Skyvva Version 2.51.9.
- With the input “2.51.9” yielding “2519”, the Agent identifies it as Skyvva Version 2.51.9.
- In the case of an incorrect input such as “abc.12”, the log will throw a NumberFormatException with the message “Invalid version format: abc.12”.
Resolved Issue: Updated Job Comment for No Data Availability in Inbound Processing via Agent Database Adapter #
During inbound processing using the Agent Database adapter, if the database contains no records for integration, the job comment displays the message: ‘Failed to process message: No input JSON objects to process.’ This issue has been resolved by updating the job comment to read ‘No data available to process.’
Resolved: Incorrect Record Count Records in Bulk Basket Size Processing with New Agent #
There is an issue when processing a Skyvva bulk basket using a New Agent with a specified Bulk Package Size. For example, if you set the Bulk Package Size to 1 record per Bulk Basket, the Bulk Basket is still created with a size of two. This issue has been resolved to ensure that the bulk basket size is correctly created from the Agent side.
Fixed the Handling of Database Tables and Columns with Hyphens #
There was an issue when the table or field name contained a hyphen, and Quoted Identifiers were not added to handle queries or insert records into this database table. This would result in the error message: ‘Incorrect syntax near ‘-‘.” This issue has been resolved by ensuring that the New Agent automatically adds Quoted Identifiers when table or column names contain hyphens.
Fixed a Few Issues with Agent Cache Postgres Database #
There are a few issues with the Agent Cache Postgres database, as follows:
- We identified that the scheduler failed to start due to the error: ‘Problem in Starting Scheduler, Error: Couldn’t store trigger ‘DEFAULT_TRG_STNG_xxxxxxxxxxxxxxx’ for ‘DEFAULT_TRG_STNG_xxxxxxxxxxxxxxx’ job:Error: value too long for type character varying(120).’ This occurs because the cron_expression column in the Agent Cache Postgres database supports only 120 characters, while our scheduler generates a longer cron expression. To resolve this, we increased the field length from 120 to 3000 characters.
- Another issue arises when the scheduler cannot start. This problem occurs only when a qrtz_triggers entry exists without a corresponding qrtz_cron_triggers. In such cases, the agent scheduler cannot create a new cron trigger and fails. To address this scenario, we implemented a check: if a qrtz_triggers exists without the corresponding qrtz_cron_triggers, we automatically remove the orphan qrtz_triggers and restart the scheduler when the user clicks ‘Start’ Scheduler. This ensures that the scheduler starts cleanly without errors.
Fixed: Disable the Save Button on Scheduler Settings While the Agent Scheduler is Running #
When the Agent Scheduler is running, you may attempt to save new settings. However, you could encounter an unexpected error after doing so while the Agent Scheduler is active. This issue has been resolved by disabling the ‘Save’ button when the Agent Scheduler is running.
Fixed ReferenceField Not Working with Inbound Agent File Adapter #
When importing files such as CSV, XML, and JSON with a reference field for a binary file, the Inbound Agent File Adapter will not work with the reference binary file. It only functions with business records. This issue has been resolved by ensuring that both business and reference files can be imported.
The flat table does not support child mapping when processed hierarchically with the Agent Database Adapter #
There is an issue when you have a flat table that stores two sObjects, for example, Account as the parent and Content Version as the child. In this case, you want to perform an inbound processing via the Agent Database Adapter. Your mapping uses a flat structure in the source field to map to a hierarchical sObject as the target. Previously, you would only receive the parent record. We have now resolved this issue, enabling child record mapping as well, so that you will receive both sObjects: Account as the parent and Content Version as the child in Salesforce.
Fixed Issue with Interface Execution for User Linked to Agent Connection of Agent xxxx Adapter #
We encountered a problem while executing the interface specified by the user. This issue arose because the Agent used the Instance ID to obtain an access token for execution as the user, which resulted in incorrect user execution of the interface within the Agent Connection linked to the adapter. To resolve this, we changed the method of obtaining the access token to use the Connection ID. Now, the new agent correctly executes the interface that links Agent xxx Adapter of Agent connection user the user with the connection link in the adapter.
Fixed Issue: New Agent ODBC Cannot Retrieve Records When the Varchar Column is Null in Records #
There is an issue with the Agent Database Adapter: the Agent ODBC 32-bit cannot retrieve records for processing when the varchar column has a null value, even though there are records available in the database. It will raise an issue in the log as: ‘java.sql.SQLException: No data found java.lang.RuntimeException: java.sql.SQLException: No data found. The issue arises from retrieving the column with the varchar as an object in ODBC. This issue has now been resolved by converting the value to a string instead of an object.
Fixed Agent Database did not get value from WHERE-Expression #
This issue has been addressed by Upsert record into the database with the V3 interface design, with passing dynamic value to WHERE-Clause; however, the WHERE-Expression set to build the query with where condition, and the get field is null, and the value is null that why an error is raised, invalidating the SQL statement.
Fixed Issue with Agent Database Not Retrieving Value and Field from WHERE Expression #
This issue has been addressed by implementing an upsert of records into the database using the V3 interface design, which allows dynamic values to be passed to the WHERE clause. However, when the WHERE expression is set to construct the query to search for records before upserting them into the database, if both the field and the value are null, an error is raised, resulting in an invalid SQL statement.
Fixed the issue with Internal Key Management on Insert Operation Mode While Inserting Records #
The problem arose when an upsert operation was attempted using the V3 interface design. If a record was not found in the database table, the system would attempt to insert the record. However, when the primary key was configured as an auto-number, and the user had set Internal Key Management in the Insert Operation Mode Properties of the Agent Database, the system did not ignore the primary key as specified in the Skyvva Mapping tool. This resulted in the error message: “Cannot insert explicit value for identity column in table ‘tablename’ when IDENTITY_INSERT is set to OFF.” This issue has been resolved by ensuring that the primary key field is ignored during the record insertion process. Users can now successfully insert records without encountering the identity column error.
Resolved Stuck Job Issue in Agent File Adapter #
An issue occurred when processing files with the Agent File Adapter, where the job could become stuck during the deletion of the processing route after execution. This has now been optimized to properly handle this scenario and prevent jobs from getting stuck.
Fix: Updated JVM Memory Configuration to Prevent New Agent Crash #
An issue occurred where the New Agent could crash during execution. To address this, we have updated the JVM configuration in the Skyvva-Agent-Service.xml file for the New Agent Service to better control memory usage. The previous setting used 70% of system RAM, which has now been changed to fixed values: -Xms512m and -Xmx700m.
