Reltio Connected Customer 360: 2020.1 Weekly Releases

This page lists bug fixes and enhancements applied as part of our Weekly Releases in Reltio Connected Customer 360: 2020.1.

Release Notes for 2020.1.15.1 Emergency Bug Fix Release - June 04, 2020

Fixes

Match IQ

Queue size was not getting reduced to zero. This issue occurred due to the incorrectly applied logic for potential matches' event processing that created additional ENTITIES_MATCHES_CHANGED events with repeated ML prefix in the CRUD queue. This issue is fixed by using the required logic correctly.

Release Notes for 2020.1.15.0 Weekly Release - June 03, 2020

What's New

Hub

Admin user was unable to reset the password for a user and this issue is now fixed. On clicking the Reset Password button, an email is sent to the user with the instructions to reset the password.

Reltio Platform

With this release, we have updated the object limits for sub-nested attributes, sub-reference attributes, and reference crosswalks per entity. For more information, see Object Size Limits

Reference Data Management

While exporting data from RDM, you can now choose to export the updateDate and updatedBy fields along with the version information of lookup types. You can choose to export in JSON format as well. For more information, see Exporting RDM Data.

Salesforce Connector

The Salesforce Connector API now supports the Synchronization API, which is used to force the synchronization process between Reltio and Salesforce for specific Reltio Objects through the synchronizing and merging processes. For more information, see Salesforce Connector API.

Match and Merge

With this release, we have added two new inspections to the dynamic Match Rule Analyzer. For more information, see Inspections for Useless and Redundant Match Groups.

D&B Connector

This operation provides a hierarchical defined structure to complete the organization hierarchy to its global ultimate account. Using the Upward Account Hierarchy feature, you can request all parent organizations (all organizations from the current organization to its global ultimate) for the current organization. For more information, see On Demand APIs

Fixes

Reltio Platform

  • In Database, while trying to perform a GET entity API call, the following error message was displayed periodically: Too many simultaneous connections (max=512).. This issue is now fixed.
  • As a mandatory check for input mapping, it is required to check if an attribute has a value or not. However, this check failed to identify if the attribute had at least one Operational Value set to true. As a result, the cleansed output was not as expected. This issue is now fixed. As an additional check, the cleanser value list for an attribute cannot be empty for mandatory attributes.

Security

A user with the ROLE_ADMIN_CUSTOMER role could not update certain roles. This issue occurred due to a restriction related to an internal resource that was part of the new role permission framework. With this release, this issue is fixed and a check is introduced to validate the changed resource in the role if no internal resource is modified

Hub

For an entity, as per an existing condition under the beforeDelete hook, if the condition is not met then the entity cannot be deleted and an error message must be displayed for the same on the Hub. It was found that the error message was not displayed on the Hub. The logs, however, showed that the error message was displayed. With this release, this issue is now fixed and the error message is displayed as expected.

Reference Data Management

  • Timestamp values like 2000-01-01T00:00:01Z were accepted by the RDM API but were rejected by the RDM UI. Therefore, incorrect validation errors were displayed in the RDM UI and canonical codes could not be updated. This issue is now fixed.
  • Unmapped lookup values were displayed on the RDM UI only for users with Edit access. This issue is now fixed and the unmapped lookup values are displayed for users with both Edit or Read access.

Life Cycle Actions

With this release, we have introduced a custom LCA for end date and start date validation with an LCA hook. As a part of this change, we have updated the LCA Jars for the end date and start date methods.

DTSS

We fixed a bug in the DTSS connections configuration that was previously incorrectly importing both the reference entity and relation for each request when the relation was part of a reference attribute between the same entity types.

Release Notes for 2020.1.14.0 Weekly Release - May 27, 2020

What's New

Workflow

We have enhanced the Potential Match Review email notification for Workflow. Upon submission for approval, the notification now indicates if the action was a merge or not a match.

Match and Merge

  • We have added two new inspections to the dynamic Match Rule Analyzer. For more information, see Inspections for Tokens
  • We have introduced a new task called ProcessRebuiltMatchTask to process any kind of matches discovered by a Spark match job. For more information, see Process Rebuilt Match Task.
  • With this release, we have introduced the truncateMatches API to clear match tables. For more information, see Clearing Match Tables.

Fixes

Salesforce Connector

When entity attributes are merged in Reltio and mapped to the Salesforce custom objects, the Salesforce Connector executes the deduplication process. This sets the deleteDate field for all attribute crosswalks except one and also removes the duplicate objects from Salesforce. It was found that during the deduplication process, the Salesforce connector did not take into account the reference attributes. This issue is now fixed.

Match and Merge

Match documents were being compared based on the list of match groups present in the first match document only. There was no check to find if the second match document contained the same match groups. Even if a match group was not found in the second document, the match document was still compared. This issue is now fixed by including the check for the second match document when comparing.

DTSS

When using the GET synchronization configuration or the manual_match APIs, the DTSS service encountered errors indicating that the service was not available. As a result, the DTSS service became unstable. This issue occurred because after migrating from the database to Google Big Table (GBT), the history of the GBT Google credentials was not added to the properties file. This issue is now fixed.

Search

The Search functionality did not display appropriate results when searched for Entity IDs. This issue occurred due to an internal optimization that led to an error when searching for numeric attributes. With this fix, search by using Entity IDs returns appropriate results.

Reltio Platform

  • It was found that the filtering for sub-nested attributes were not getting applied when filtering nested attributes. This issue is now fixed and sub-nested attributes are getting filtered.
  • In this release, the members, members.types, and members.{{Type}}.id parameters are added to the entities/{{entity_id}}/_interactions endpoint to support the filter by operation. For more information, see Get Interactions for an Entity.
  • For some specific cases related to verticals, an error code was not displayed for abstract classes. This issue is now fixed and the error is displayed as a warning.

Release Notes for 2020.1.13.0 Weekly Release - May 20, 2020

What's New

Console

We have introduced the forceIgnoreInStreaming option in the Tenant Management application within the re-index task. This option forces the re-index task to ignore streaming events and blocks the generation of all events during the re-indexing process. This option was already available in the API and has now been added in the Tenant Management application. For more information, see Creating a Reindex Data Job.

Reltio Platform

We have added new object limits for sub-nested attributes, sub-reference attributes, and reference crosswalks per entity. For more information, see Object Size Limits.

Salesforce Connector

You can now synchronize a relation in the Reltio Platform to Salesforce as an object. For more information, see Relation Synchronization from Reltio Platform to Salesforce.

Reltio Data Science

We have added matchRule URI data to the view_merges_sources view of Analytics and the merges table in Data Science. This enhancement allows you to uniquely identify match rules (by URI) in addition to the labels that were responsible for the merge. For more information, see Defining Datasets with Merges.

Reference Data Management

MDM cache is enhanced with an additional option to enable faster transcoding of lookups in memory. For more information, see RDM Setup and Migration.

D&B Connector

Prior to this release, the CMPELK dataset was only supported in Real-time mode. Now, it is also supported in the Batch mode. For more information, see Create D&B Batch Job and D&B Connector.

Fixes

Life Cycle Actions

LCA request for the afterSave hook was taking more than one minute to process one entity. It was found that the internal token's validity was set to one minute. As a result, the REST API calls made from the LCA Service failed with the 401 error. This issue is fixed and the internal token's validity is increased to three minutes by default.

Hub

In multiple cases, users without the permission to merge could see the Create Relation button on the Potential Matches screen even if the button was not enabled for them. This issue is now fixed.

Salesforce Connector

In Salesforce, the equals(<SomeAttributeValue>, true) filter was always returning False and therefore the filter could not be used as expected. This issue is now fixed.

Reltio Platform

  • The Relation Event Scan API did not work as expected for the deleted events. It was found that the relation scan endpoint was not implemented on tenant index in the correct way. This issue is now fixed and the endpoint is correctly implemented.
  • The status of Data Import jobs was found to be stuck as JOB_PENDING for a long time. It was found that multiple jobs with a large number of nodes failed on the database cluster, but continued to show the PROCESSING status. Therefore, such jobs stopped/prevented the processing of other jobs in the queue. This issue is now fixed by canceling all such problematic jobs and then re-invoking all those jobs that were canceled during the investigation of the problematic jobs.
  • When an auto-generated field uses the "generateIfEmpty": true and "generateIfNotEmpty": false settings, the platform does not generate values if the user provides the values. If the user does not provide any value, then the platform creates auto-generated values. However, it was observed that in certain cases, if the user provided duplicate values, the auto-generated fields were not merged. This issue is now fixed by checking that the auto-generated field values are applied/used correctly, and then, removing the duplicate values if found.

Release Notes for 2020.1.12.0 Weekly Release - May 14, 2020

What's New

Reltio Platform

With this release, we have de-activated the use of the force flag that was available in the re-index task. The purpose of this flag was to ensure consistency when versioning was not available for entities. Now, as we have versioning available for entities, the purpose of this flag is already achieved. Additionally, we wanted to provide you with some additional time so that you can remove this flag from your re-index calls. This change does not break any existing functionality. Note that if you continue to pass the force flag in the re-index task, behind the scenes, it will have no effect.

Fixes

DTSS

While performing a manual import from a Data Tenant (DT) to a Customer Tenant (CT), the updateDate attribute was not updated to the current date and time. As a result, a mismatch in update date was found between DT and CT after the import. This issue is now fixed by setting the updateDate attribute to the current date and time before performing a manual import from DT to CT.

Hub

A lot of time was utilized for processing the historic URI requests and that hung the History tab. This issue is fixed by limiting the number of entities to 1000 when processing the historic URI request.

Salesforce Connector

While reading the content from the Reltio_Sync__c table, the status check of Salesforce Connector was not retrieving more than 2000 records. This issue is now fixed by changing the pattern of reading the Reltio_Sync__c table to read only the number of event types.

Reltio Platform

  • The activities/_scan endpoint was taking more than the expected time for large data extractions. This issue is fixed and the performance of the endpoint is improved. In addition to this, an option to disable the creation of object labels in the activities/_scan request is added to the tenant configuration.
  • When completing reindexing in the background, Reltio switches the existing alias to point to the new index from the old index. At the same time, there could be other indexing events that are being processed. Switching the alias to point to the new index is automatic with the tenant index. Records are indexed by first resolving the alias and then pushing the record into the index pointed to by the alias. Though the alias switched to the new index, the tenant index pushed the record to an old index. As a result, the IndexNotFound exception was reported. In this case, the Reltio platform created an empty index that resulted in a large number of entities not found. As a fix to this issue, before creating a new index, the platform verifies whether the alias already exists and creates a new index only if the alias does not exist. If the alias already exists, the record is indexed again and does not result in the IndexNotFound exception.
  • An index name includes the date obtained from the timestamp. However, it was found that the index names for Activity Log and Interactions contained a reference of a custom timestamp that was created from an object. As a result, the timestamp was found to be broken and this led to the creation of a large number of indices. This issue is now fixed by considering the timestamp associated with object creation time.

Console

  • The pattern for subscription name of CRUD and MATCH queue is -crud-/-match-. However, this pattern was not considered for account subscription name for several Google Cloud Platform (GCP) environment _crud-/_match- queues. As a result, the information related to queues was not available on the Tenant Management application. This issue is now fixed and the account subscription name is read from the system configuration.
  • Multiple customers with SSO authentication could not access the External Match UI application from Console and received the 400 error. These errors were caused because the header contained cookies with very large tokens generated by the SSO authorization service. This issue is now fixed by removing the cookies before sending the UI request.

Release Notes for 2020.1.11.0 Weekly Release - May 06, 2020

Fixes

DTSS

When using the GET synchronization configuration or the manual_match APIs, the DTSS service encountered errors indicating that the service was not available. As a result the DTSS service became unstable. This issue occurred because after migrating from database to Google Big Table (GBT), the history of the GBT Google credentials were not added to the properties file. This issue is now fixed.

Reltio Platform

  • When preparing an object for indexing in tenants with Operational Value (OV) Calculator V2, OV was calculated multiple times. For every nested or reference attribute, the OV was recalculated. It became more time-consuming in the case of group OV strategy calculation, as OVs were recalculated for each survivorship group. This issue is now resolved and OVs are calculated once for each survivorship group.
  • An internal synchronization process produced unnecessary and/or incorrect synchronization of match documents when metadata tenant configuration was updated while the synchronization process was running. This issue produced a large number of events. This issue is now fixed. After the release of this fix, you may see a one-time spike of events and this process may resync some entities that were earlier synced incorrectly.

Release Notes for 2020.1.10.0 Weekly Release - Apr 29, 2020

What's New

Hub

Earlier, the Reset Password functionality would send the password reset link through emails to the users who raised the password reset request. This was analyzed to be a vulnerable security practice. Now, the Reset Password functionality sends a verification code that can be used to reset the password. For more information, see Reset Your Password.

Reltio Platform

The Password Reset functionality is modified. As part of this change, the Send Reset Password Link API (POST {oauth_uri}/users/_sendResetPasswordLink) is deprecated and the Request for Verification Code API (POST {oauth_uri}/users/sendVerificationCode)is introduced. Now when a user requests for resetting the password, a verification code is sent instead of the password reset link.

Workflow

  • We have added a spinner to indicate the progress of an approved/rejected Data Change Request (DCR) action.
  • We have added the support for rejection of sub-attributes of a newly created nested attribute. For more information, see About Data Change Request.

Console

  • We now provide the ability to cancel Data Loader Jobs using the DL Jobs API. Once the Job is canceled, it can not be resumed.
  • We have introduced a new role in Data Loader to manage accessibility to Data Loader resources. For more information, see Roles and Permissions

Match and Merge

  • We have introduced a new parameter called entityTypes for RebuildMatchTable task version 2. For more information, see Rebuild Match Table Task Version 2.
  • We have added support for distributed mode for PotentialMatchesReindexTask task. Now, you can run the PotentialMatchesReindexTask task in multiple parts, in parallel, to enhance the performance of the job. For more information, see Potential Matches Reindex Task
  • We have added support for distributed mode for the MergeAutoMatchesTask task. Now, you can run the MergeAutoMatchesTask task in multiple parts, in parallel, to enhance the performance of the job. For more information, see Merge Auto Matches Task.
  • We have added a new calculator named CountFrequency to the Match Rule Analyzer. Using this calculator, you can retrieve the token counts for the entities.

Fixes

Reference Data Management

A lookup attribute with value="lookupRawValue" could not be merged with a lookup attribute with value="lookupCode" after enabling the "resolveLookupCode" option. As a result, lookup attributes were duplicated. This issue is fixed by ensuring that the lookups are resolved during cumulative updates.

Reltio Platform

  • Before this fix, sub-nested attributes were not inherited from the lower level configurations. With this fix, Reltio now inherits sub-nested attributes from all the configuration levels even if they are not set in the metadata configuration or are set partially. For more information, see Reltio Information Model and Configuration Concepts.
  • For entities with the ID generators enabled and duplicate attribute values sent into the body of the request, the duplicate values were not collapsed. This issue is now fixed and the already existing duplicate attributes are collapsed using the removeAttributeDuplicates periodic task.
  • In multiple cases, entities were not readable. This issue occurred because the user permissions were not generated properly in case of Single Sign On (SSO), if the user had the groups and tenantPerRoleSecurity options turned on. This issue is now fixed and the entities are readable.

Release Notes for 2020.1.9.0 Weekly Release - Apr 22, 2020

What's New

Hub

Reltio now supports basic filtering on interactions related to an individual profile. For more information, see Interactions Search.

Salesforce Connector

  • You can now sync the entityURI system attribute from the Reltio platform to Salesforce. For more information, see Mapping Functions.
  • In the Salesforce Connector Mapping Configuration, you can now use the AND or OR condition in the filter expression. For more information, see Data Filtering.
  • You can now synchronize relationships from the Reltio platform to Salesforce. For more information, see Relation Synchronization from Reltio Platform to Salesforce.

Fixes

DTSS

DTSS task created new entities in the Customer Tenant (CT) instead of merging them with the existing entities. The DT_DEA crosswalk was not used as expected and as a result, Data Stewards had to manually process the matching profiles. This issue occurred as the cleanseConfig configuration crosswalk type was not properly filtered. This issue is now fixed by ensuring that the cleanseConfig crosswalk type is properly filtered.

Reltio Platform

  • In some scenarios, incorrect operational values were selected for nested attributes when SRC_SYS was set in the survivorshipAdvancedBehavior configuration. As a result of this, the OV calculation step that considered multiple crosswalks, was skipped due to optimization. This issue is now fixed by ensuring that the step is not missed.
  • Reindexing jobs remained in the SCHEDULED status even after the tasks were initiated. This issue occurred as the platform failed to estimate the number of nodes that a cluster needed to run all active tasks in parallel for a tenant. This issue is now fixed by correcting the estimation logic.

Console

  • Multiple customers with SSO authentication could not access the Data Loader application from Console and received the 400 error. These errors were caused because the header contained cookies with very large tokens generated by the SSO authorization service. This issue is now fixed by removing the cookies before sending the UI request.
  • Multiple customers with SSO authentication could not access the Data Modeler application from Console and received the 400 error. These errors were caused because the header contained cookies with very large tokens generated by the SSO authorization service. This issue is now fixed by removing the cookies before sending the UI request.
  • Multiple customers with SSO authentication could not access the SSO Configuration application from Console and received the 400 error. These errors were caused because the header contained cookies with very large tokens generated by the SSO authorization service. This issue is now fixed by removing the cookies before sending the UI request.

Release Notes for 2020.1.8.0 Weekly Release - Apr 15, 2020

What's New

Reltio Platform

To simplify the configuration of immutable reference attributes, few additional abilities have been added. To know more about the immutable and immutableExceptForSources properties, see Immutable Reference Attributes

Fixes

DTSS

Single attribute view did not show values for a DTSS entity. This issue is fixed and the Hub now shows the attribute values of entity objects from DTSS.

Match and Merge

After running the RebuildMatchTable, PostRebuildMatchTableCheck, and MemorySafeCassandraMatchConsistency tasks, inconsistencies were found in the match documents. This issue occurred due to the improper synchronization in the matching storage for entities where reference attributes reached the maximum limit (maxAttributeValues). This issue is now fixed by correcting the algorithm to tackle the synchronization between storages.

Reltio Platform

  • While adding a new address to an account and saving the change, an internal application error message appeared and address was not saved. This issue is fixed by ensuring that the IGNORE operation is always performed without creating a new crosswalk.
  • The nextEntry parameter of the Connections API was not working as expected when specified together with the filter parameter. When both the filter and the nextEntry properties were applied to filter out the first level of records for a particular entity (parent entity), it was found that the records (child entities) of not just that parent entity but other entities were also displayed. This issue is fixed and the filter and the nextEntry properties now display records (child entities) of the parent entity only.
  • The unmergeAllContributor API was not working as expected. The entities were unmerged and again merged back when the unmergeAllContributor API call was made. As a result, entities could not be successfully unmerged. This issue is now fixed by removing the phantom entity to prevent merging by Crosswalk.
  • While defining survivorship rules for attributes, it was observed that SurvivingSourceID was giving the correct value but SurvivingSource gave multiple values instead of one value. This issue occurred as the OtherAttributeWinnerCrosswalk strategy failed to check whether the current attribute was the parent of the primary attribute. This issue is now fixed.
  • Cassandra export tasks were causing performance issues. This issue occurred due to the increase in the number of threads after an internal optimization, which created excessive load on Cassandra. This issue is now fixed by decreasing the default value of the threads by a factor of two.
  • There was an issue with the conversion of the time and timestamp fields. In Reltio platform, time and timestamp are considered the same as both contain time and date. However, in BigQuery time refers to the actual time without any date. This issue is fixed by making appropriate changes in data transfer of the time and timestamp attributes from Reltio platform to BigQuery.

Export

  • Incorrect OV values were extracted when performing Spark based export. This issue occurred because Spark export did not consider user roles when calculating OV values (for selecting the appropriate survivorship strategy). This issue is now fixed.
  • Export task failed and also displayed an error message that the specified upload did not exist. S3 multipart upload allows using a part number from 1 to 10,000. The task failed because the part size was 5 MB and the limit of 10,000 part number was exceeded. This issue is now fixed by calculating and increasing the part size accordingly.

Hub

  • The older edit mode of Search had some issue. You could click the Search icon and enter any text value to filter the Relation View results. After removing the added text values and on clicking Enter, the older Relation Views would appear. However, this was not how the Hub was behaving. The Hub sent an incorrect request after you clicked Enter. This issue is fixed as the Hub now sends the correct request and you can see all the available relations.
  • History of an entity was not available on the Hub and also while making API calls. This issue occurred because the mandatory taskAssignee property was missing during the deserialization (conversion from JSON to Java Object) of business process tasks related to the entity in history. This issue is now fixed by making the taskAssignee property optional.
  • While updating the lookup attribute value from the Hub, the lookupRaw value was updated with the lookupCode value. This issue is fixed by correctly restoring the raw values for the lookup attributes.
  • In case of large error messages, the complete error message could not be read and the message box could not be closed. In addition to this, error message that appeared after performing the Merge All action was not visible. This issue is fixed by setting the max-width property on the error message box to keep the message smaller than the screen size. Also, the Merge All action now shows the errors, if any.

Release Notes for 2020.1.7.0 Weekly Release - Apr 08, 2020

What's New

Hub

Previously, only users who created saved searches could delete them. Now, the tenant administrator can delete saved searches created by any user.

Reltio Platform

The responses of the Job Management API calls have been improved to include the duration parameter when a subtask or job is completed. This parameters allows you to estimate the time required to complete the subtask or job.

Export

Added an option to merge multiple export output files into a single file for download by a user. For more information, see Exporting Entities and Relationships.

Console

  • Reltio now ensures a consistent date and time format for export jobs in the Export UI of Console. All timestamps are shown in the user's local time zone. For more information, see Downloading Exported Output Files.
  • Reorganized layout of the SSO SAML configuration page in Console to accommodate user group settings. This allows the user group SAML configuration to be updated online. Previously API calls were required to update the SAML user group configuration. For more information, see SAML SSO Configuration.

Fixes

Console

The mapping processor was unable to construct JSON records if the attribute values were missing in the CSV file. As a result, Data Loader was getting stuck multiple times. This issue is fixed by improving the code to remove the reference entity crosswalk if the attribute types are missing in the CSV file.

Match and Merge

The Distinct Words Comparator was not working in a symmetrical manner. This means that the result of comparing value1, value2 and value2, value1 was not found to be equal. The comparator is now fixed to behave in a symmetrical manner.

Reltio Platform

  • An issue was observed while extracting data from a downloaded file as the data could not be successfully unzipped. This issue is fixed by correcting the logic associated with the order of the zipped file parts.
  • Periodic tasks were completed without processing an entire dataset due to a throttling issue. This occurred because the synchronization between internal processes had an issue that incorrectly updated the task status. As a result, the status of the periodic tasks was set to COMPLETED and the task stopped processing the dataset. This issue is now fixed.

Release Notes for 2020.1.6.0 Weekly Release - Apr 01, 2020

What's New

Match and Merge

A new consistency task named CassandraPotentialMatchesConsistencyTask is introduced to compare entities and potential matches. For more information, see Cassandra Potential Matches Consistency Task

Fixes

Reltio Platform

There is a process for merging entities on-the-fly during data load. A similar process for merging relations was having some issues. A mismatch was found in the code base for entity and relation merges on-the-fly during data load. This led to the failure of data load for large batches. We have fixed this error internally and given the fact that both entities and relations follow the same process, the data load must complete successfully.

DTSS

All the subscribed entities failed to sync when using the manual_sync task. Data collected on behalf of multiple duplicate crosswalks created multiple duplicate crosswalks and resulted in this issue. Prior to this fix, DTSS could retry and delete a maximum of seven duplicate crosswalks only. To fix this issue, the DTSS service is updated and a new property named multiple.contributors.error.max.retries is added to allow custom number of retries to delete all the duplicate crosswalks. In addition, the logging capability is improved.

Hub

When using the older Profile mode, the search did not work as expected in Relations View as the connection requests sent from the older UI had the incorrect filter. This issue is fixed as the older UI now sends the correct connection request.

Console

  • For Console modules like Tenant Management and Data Loader, the Completed Job page was visible but an error occurred on the New Job page. This issue is now fixed by excluding the access and refresh tokens from the API calls.
  • The SSO Login screen was hanging up frequently. This issue occurred due to the implementation of an older loading component that caused infinite loops for the SSO Provider if the skipReltioLogin parameter was set to false. A new loading component is introduced to fix this issue.

Workflow

  • We have fixed a bug that was preventing all changes to be applied as requested in a Data Change Request (DCR) if the DCR contained multiple changes like:
    • Entity updates from crosswalks
    • Remove relations
    • Add relations

    All changes are now applied correctly.

  • The UserTask in Workflow was created with a day's delay. This issue occurred as the activiti:async parameter was set to true in the custom Workflow process. Therefore, the Business Process Model and Notation (BPMN) process was used in the asynchronous mode. To avoid this delay, it is recommended that you do not use the asynchronous mode in the BPMN process.

Release Notes for 2020.1.5.0 Weekly Release - Mar 25, 2020

What's New

Export

We optimized the writing of the Export output file. This improves the performance of the Export service when Native Export is used (when the primary data storage is used as the Export data source). For more information, see the following pages:

Fixes

Console

Customers with the skipReltioLogin parameter set to false in the UI configuration were unable to login through SSO provider by clicking the Login with SSO UI button. This issue is fixed by correcting the client-side service to return the correct SSO URL.

Export

When you performed a CSV export, all attributes except uri, updatedTime, and createdTime attributes were extracted. This issue occurred as these attributes could not be selected in the CSV file. The issue is now fixed.

Search

When you selected the Search by operational value only checkbox and performed a search by using the ID attribute value, Search did not return the expected entity. This issue occurred because an earlier fix related to optimizing the search engine was not applied to nested attributes while indexing. As a result, Operational Value (OV) of all the simple attributes were recalculated and overwritten for all the survivorship groups. This issue is now resolved by applying the earlier fix for nested attributes also to ensure that the simple attributes are not recalculated (overwritten).

Salesforce Connector

When importing records using Search Before Create (SBC) from MDM Customer Tenant, Salesforce crosswalk data was not added to the records. As a result, duplicates were created in MDM for certain cases. This issue is now fixed by adding the Salesforce crosswalk while importing records.

Reference Data Management

The RDM UI was unable to navigate to the tenant selection page and was instead redirected to the Login page. This issue is fixed by first selecting the tenant and then authorizing the user.

Reltio Platform

  • After running the reindex task with the separate index tasks enabled, entities could not be searched using URI. This issue occurred as the match rules cleanse dictionaries contained invalid patterns. For example, https://s3.amazonaws.com/reltio.match.test/hp/hp_sr_org_20181126.txt. . This issue is fixed by changing the invalid pattern as follows:
    • "\=>\" to "\=>\\"
    • "\bb\.parque capital(cba\b" to "\bb\.parque capital(cba)\b"
  • When a relation was created using an API request that had the URIs of the start and end entity, the relation was getting created but the startObjectCrosswalks and endObjectCrosswalks fields were not populated for reference attributes. This occurred because a relation is defined as a relation between crosswalks and not between entities. Therefore, when an API request contained the start/end entity URIs, the system was unable to identify the entity crosswalks that had to be connected for the relation (start/end entity URIs). As a result, when the slice service tried to calculate the entity slice by ignoring an attribute, the reference attribute was missing because the relation was not connected to any entity crosswalk. Now, the behavior of the slice service is changed to ensure that in case of a cumulative update, if the start/end object crosswalks from the reference attribute are empty, then all the relation crosswalks are used.
  • Special characters in S3 input path led to load post interaction failureThis issue occurred because the job was configured to run on 's3n' and also because of the S3 input path to the file that contained special characters like brackets, quotes, and so on. You are advised to use 's3a' to configure the jobs if there are special characters in the S3 input path to the file.

Release Notes for 2020.1.4.0 Weekly Release - Mar 19, 2020

What's New

Workflow

We fixed a bug in our change request endpoint so that now it returns the correct reference entity URI. Previously, it was returning an incorrect reference entity URI. To fix this bug and maintain backward compatibility, we had to introduce the new parameter called showObjectsMetadata, which needs to be set to true. The showObjectsMetadata parameter works exactly in the same way as the showObjectsInfo parameter with the exception that it returns the correct reference entity URI. Going forward, the showObjectsMetadata parameter must be used when using partial approval for Data Change Request (DCR) endpoints. For more information, see Get Data Change Request and Get Several Data Change Requests.

Fixes

Workflow

If an incorrect URI was generated for a reference entity while creating a DCR, the label was missing for affiliation details. This issue was caused by the incorrect URI generated for the reference entity as part of the DCR. The entity with the incorrect URI could not be found and that led to an empty label. This issue is fixed as reference entities are now resolved with the valid URIs and the labels from the referenced entities are shown correctly.

Reltio Platform

  • Nested attributes were merged incorrectly when multiple updates were applied. The issue is fixed by correctly collapsing the nested attributes.
  • Address Type did not display the lookup code and lookup raw value when an entity was retrieved through an API. This issue occurred as lookup dependencies worked incorrectly if the parent attribute was defined after the dependent attribute in the business configuration. This issue is now fixed my making the resolution of the dependent lookups independent of how the attributes are defined in the business configuration.
  • Crosswalks deleted using APIs were still available even though the activity log and history showed the deletion details. This issue is fixed by implementing the retry mechanism for concurrent updates made to objects. If the retry count exceeds three, then the TOO_MANY_CONCURRENT_REQUESTS_FOR_OBJECT exception is displayed. You can either repeat the request later or change the logic to avoid multiple concurrent operations for the objects mentioned in the exception.
  • Concurrent changes made to an object led to the failure of the merge operation. This issue is fixed by implementing the retry mechanism for updates made to objects during the merge operation. If the retry count exceeds five, then the TOO_MANY_CONCURRENT_REQUESTS_FOR_OBJECT exception is displayed. You can either repeat the request later or change the logic to avoid multiple concurrent operations for the objects mentioned in the exception.

Hub

  • Inconsistent behavior was observed when creating a profile. By default, a message appeared on the screen for few seconds mentioning You do not have access to edit this attribute and then disappeared before moving ahead with the profile creation. This issue is now fixed by hiding the message that appeared by default.
  • After saving the changes made to the Profile page, the page could still be edited in the Viewing mode. This issue occurred as the dependent lookups were not processed correctly and the UI reported an exception. Now, this issue is fixed by making a few changes to the lookup processing code.
  • After making changes through a Data Change Request (DCR), some entities were not visible in the Viewing mode. This issue occurred as the dependent lookups were not processed correctly and the UI reported an exception. Now, this issue is fixed by making a few changes to the lookup processing code.

Release Notes for 2020.1.3.0 Weekly Release - Mar 12, 2020

What's New

Reltio Platform

  • A new OV strategy called GROUPS is added. It is similar to STATIC, but additionally it indexes all survivorship groups and the search is performed using a survivorship group that is associated with a user role in the business configuration.
  • Platform is enhanced to detect and reindex the changed entities. This enhancement modifies the Create new job screen for the Reindex job in Tenant Management within Console. As part of the entity reindex optimization, the platform detects and re-indexes only the changed entities. Therefore, the Ignore version conflicts option that allowed you to reindex the entities even if the entities have not changed is removed. For more information, see Creating a Reindex Data Job.

Console

  • The Tenant Management application is enhanced to display the tenant information on all screens. Tenant information is available in the read-only mode indicating that you cannot change the tenant during a process/task/job.
  • The External Match application is enhanced to display the tenant information on all screens. Tenant information is available in the read-only mode indicating that you cannot change the tenant during a process/task/job.
  • The Export application is enhanced to display the tenant information on all screens. Tenant information is available in the read-only mode indicating that you cannot change the tenant during a process/task/job.

Fixes

DTSS

When a crosswalk date was set as a future date for an entity in DT, after downloading to CT, the same future date was not reflected in the CT. Now, when a crosswalk date is set as a future date for an entity in DT, after downloading to CT, the same future date is reflected in the CT.

Match and Merge

The _matches API call did not display the expected results. The changes to discover errors in the platform tasks engine led to an implicit change of internal structure to store information about external match limits. This led to errors while reading the existing limits.As a workaround, you can use the /entities/_matches/_resetExternalMatchLimit API call to clear obsolete information.

Workflow

When a user approved a DCR that contained an entity and a relation ("hasaddress"), the relation was not created but a separate address entity was created. This issue is fixed now.

Reltio Platform

  • When trying to group survivorship rules as per Country and Source, appropriate results were not displayed. However, on removing the Country filter, results were displayed. The issue occurred as the search engine performed the search based on an invalid survivorship group that led to no results. With this release, this issue is resolved by fixing the search engine to use the survivorship group in the correct manner.
  • The Operational Value (OV) flags were updated with incorrect information when the data was exported.

    This issue occurred because as part of performance improvements, the marshaling stage was moved to a separate thread, but the thread context was not copied to the new one. This resulted in missing user roles in the new thread. This issue is fixed and the exported data displays the correct operational values.

Export

  • Relationship details were not exported properly. This issue happened because invalid query parameter or filter values were passed to the Export Relation API. This issue is fixed by ensuring that the filter values are ignored while exporting relations.
  • The results of Export APIs were different based on the tasks that were triggered by the smart export logic. If a ‘native’ task was triggered, the exploded parameter was considered and the results would contain "exploded" objects, as expected. However, if a ‘spark’ task was triggered, the exploded parameter was ignored. This issue occurred because the RIQ Export job did not support the exploded parameter.

    This issue is fixed by ensuring that the RIQ Export job supports the exploded parameter.

  • The users were unable to extract data from the tenant while using client credentials. This happened because the Export job did not support client credentials. This issue is fixed by ensuring that the client credentials are supported for running an export job.

Release Notes for 2020.1.2.0 Weekly Release - Mar 04, 2020

What's New

Security

User Role Assignment by Tenant ID support in Metadata Security. This metadata security enhancement allows the same role names and definitions to be reused across tenants. Users can be assigned appropriate roles at the tenant level. This allows, for example, a developer to be assigned full access to development tenants and read-only access to test and production tenants. For more information, see User Role Assignment by Tenant ID.

Console

Dataloader UI is updated to display the tenant information on wizard steps. Tenant information in read-only mode indicates that the user cannot change the tenant in the wizard steps.

Fixes

Data Cleansing

When the international country code (for example, +1 for US and +61 for Australia and so on) of a country is shared over multiple regions, Reltio's phone cleanser matches the phone number (excluding the country code) with the region-specific phone standards. However, when the country code of a particular phone number did not match any region-specific phone standards, then an incorrect country was returned. This issue is fixed. Now, if a particular country code does not match a region-specific phone standard then the international country code (main region) is considered.

Export

While exporting the entire HCP data using spark based export, it was found that multiple fields containing lookupCode and the lookupCode value in the value field did not exist. As a result, incorrect HCP data was exported for the HCPs. This issue occurred as the lookup count had a limit and if the limit was exceeded, then the lookups were not resolved. This issue is now fixed by increasing the limit of lookups to 50,000.

Workflow

The request to GET https://prod-usg-workflowui.reltio.com/external/proxy/configuration failed with the 400 error and as a result, users could not access Workflow Inbox. As per an existing functionality, access and refresh tokens can contain 400 symbols only. In this case, the header information exceeded this limit. To resolve this issue, some internal changes have been made. However, Reltio still recommends you to generate tokens that are within the 400 symbols size limitation.

Reltio Data Science

RIQ match jobs (MatchTask and GenerateMatchStructuresTask) failed to request match cleanse dictionary data from the server if the request body was not empty. This issue occurred as the POST /cleanseDictionary request body could not be parsed into a specific structure. This issue is now fixed by introducing a generic map to read the request body.

Hub

Changes made to the hierarchy tree from the Graph view were not visible until the page was refreshed. This issue occurred as the hierarchy tree layout on the graph view failed to immediately update the graph when the tree was changed. This issue is now fixed.

Reltio Platform

  • Bulk operation could not be performed while working with boolean sub-attributes. While performing a bulk operation, it was found that even after setting the value for the boolean sub-attributes to Yes and clicking Next to confirm, the default value for the entities after the bulk operation appeared as No. This issue is now fixed.
  • Verticals with issues in attribute URI had duplicate sub-nested attributes after inheritance. This issue is now fixed by improving the platform to handle attribute URIs with the missing entity type.

Release Notes for 2020.1.1.0 Weekly Release - Feb 27, 2020

What's New

Match and Merge

An additional task called PostRebuildMatchTableCheckTask is executed to check and ensure that all match documents are processed and none of the documents are missed. For more details, see Rebuild Match Table Task.

Fixes

Export

Multiple records with no loser entity record appeared in the export result even though the merges met the filter criteria. This issue occurred because the loser entities that met the filter criteria were removed if the sub-mergers of these losers were filtered out. In addition, some records appeared in the export result even if no merge took place. These issues are now fixed. Records without any merged entities are no longer included in the export results. Also, in case of sub-mergers of losers not meeting the filter criteria, the losers are not removed.

Workflow

If two Data Change Requests (DCRs) containing requests to create two entities with default crosswalks were approved sequentially and if both entities were matched by an automatic rule, then merge-on-the-fly took place. However, no information about the reason for merge (merge reason and match rule) was available. This issue is now fixed as all information about the merge is available.

Match and Merge

While running the RebuildMatchTable job, multiple CopyManualAutoMatchesTasks tasks failed because the configuration was not updated. The job now continuously checks whether the configuration is updated with a timeout (10 retries with a one-minute interval).

Console

When downloading the configuration from Data Modeler, the tenant ID was not populated in the saved L3 file. This issue is now fixed.

Salesforce Connector

Bulk updates made on Salesforce using create/update/delete requests lasted up to 30 minutes and caused sync failure. This issue is now fixed.

Reltio Platform

  • For an initial data load performed by setting resolveLookupCode=false, the Address Type associated with Address Type Code were not resolved properly. As a result, duplicate lookup attributes were displayed. This issue was fixed by running RemoveAttributeDuplicatesTask and applying resolveLookupCode=true during the initial data load. In addition, an issue related to lookup attributes not being collapsed on the fly is also fixed for override/merge based on resolveLookupCode options.
  • Zendesk password could not be reset due to reCAPTCHA error. This issue occurred becasue the old captcha keys had accidentally expired. With the newly generated captcha keys this issue is now fixed.

Release Notes for 2020.1.0.1 Emergency Bug Fix Release - Feb 20, 2020

Fixes

Reltio Platform

SSO tenant login displayed a gateway timeout error. SSO did not return the refresh token and therefore, the browser cookies had to be cleared for every login. This issue is fixed by handling the error for the refresh tokens properly.

Console

Changes made through UI Modeler are now reflected in Reltio Hub.

Release Notes for 2020.1 GA Release - Feb 16, 2020

Fixes

Life Cycle Actions

The URIs Mapping Service (a special LCA which provides information about merges and unmerges to the Collaboration Service through SQS queue) faced multiple issues. One of them was about the Reltio comments that disappeared after performing a merge operation. This issue is fixed. In addition to this, the URIs Mapping Service configuration file is introduced in S3.

Salesforce Connector

Incorrect record type was set for objects that were created by importing entities from the Search Before Create (SBC) page. This issue is fixed by using the SF Objects + record type id parameters in the SBC API to find the appropriate mapping.

Reltio Platform

  • We optimized writing export output files by eliminating writing files to local disks (files are now written directly to the destination bucket).
  • When reindexing the database, a checkpoint is maintained in such a way that upon failure, reindexing can resume from that checkpoint. Previously, a new checkpoint was only made for entities for which the updatedSince parameter was updated. This could lead to more records being reprocessed than required. The issue has been addressed by updating the checkpoint more frequently.

Export

  • We now support bi-directional reference attributes in Export version 2.
  • Start and end object crosswalks are now available in Export version 2.