Thank you for your feedback.
Form temporarily unavailable. Please try again or contact to submit your comments.

New York Patch 3

Log in to subscribe to topics and get notified when content changes.

New York Patch 3

The New York Patch 3 release contains important problem fixes.

New York Patch 3 was released on November 7, 2019.
Build date: 11-03-2019_0756
Build tag: glide-newyork-06-26-2019__patch3-10-23-2019
Important: For more information about how to upgrade an instance, see Upgrade to New York.

For more information about the release cycle, see the ServiceNow Release Cycle. For a downloadable, sortable version of New York fixed problems, see KB0748891.

Note: This version is approved for FedRAMP. The following products are not approved for FedRAMP in New York: Benchmarks, Document Viewer, Instance Data Replication, and ServiceNow mobile.


New York Patch 3 includes 143 PRBs fixed in various problem categories. The chart below shows the top 10 problem categories.

Figure 1. Top 10 problem categories
Fixed issues grouped by problem categories bar chart

Security-related fixes

New York Patch 3 includes fixes for security-related problems that affected certain ServiceNow® applications and the Now Platform®. We recommend that customers upgrade to this release for the most secure and up-to-date features. For more details on security problems fixed in New York Patch 3, refer to KB0784250.

Notable fixes

The following problems and their fixes are ordered by potential impact to customers, starting with the most significant fixes.
Problem Short description Description Steps to reproduce



Scheduled upgrade jobs are stuck on instances waiting on SOAP response and there is no socket timeout The scheduled upgrade job runs and creates a connection to the upgrade server to confirm if it needs to run an upgrade. In this process, the connection is stuck and the job is stuck running.




The email 'New certificate(s) added for the IDP(s)' is sent every 30 minutes after upgrading to New York with SSO enabled After upgrading to New York, whenever the job 'Refresh MultiSSO IDP Metadata' runs (by default, every 30 minutes), a notification email is sent with the subject 'New certificate(s) added for the IDP(s)'.
  1. Upgrade the instance to New York or provision a New York instance.
  2. Set up single sign-on.

Notice that once the job 'Refresh MultiSSO IDP Metadata' runs, the email is created with the subject 'New certificate(s) added for the IDP(s)'.




Some related lists are not showing language translations in the tab names Not all related lists associated with projects are translated when the language is changed.

Refer to the listed KB article for details.

Core Platform


Semaphores become stuck trying to check if the glide_user _activity value is Base64 encoded Semaphores become stuck trying to check if the glide_user_activity value is Base64 encoded. This can cause semaphore exhaustion and/or HTTP-429 error messages if there are enough stuck semaphores.

Appsec – Platform Dev



An error occurs when updating /inserting/deleting a user record The following error occurs when updating/inserting/ deleting a user record: 'org.mozilla.javascript. EcmaError: Cannot set property "harc_curr_val" of undefined to "No Default Credentials Found"'.

Refer to the listed KB article for details.



For change risk assessment, the calculation of the normalized value is incorrect For change risk assessment, the normalized value should be the sum of actual values multiplied by weighted values. However, currently the calculation is incorrect.

All other fixes

Problem Short description Description Steps to reproduce

Agent Workspace


Related lists are not shown on the form when a record is opened through a reference field
  1. Add a new reference field to the incident table, and reference the alm_asset table.
  2. Configure the incident form in the Workspace view so that the new field is on there.
  3. Navigate to the alm_asset table and open any record.
  4. Configure the form layout for the Workspace view by adding a few fields.
  5. Configure the related lists on the form from Step 4 and add some other lists.
  6. Open /workspace and open any incident record.
  7. Add any record to the reference field created in Step 1 and save.
  8. Click the reference icon for the reference field to open the asset.

Notice that the related lists are not shown on the form.

Agent Workspace


During the plugin loading stage in the upgrade, Agent Assist is not working
  1. Start the upgrade process.
  2. During the plugin loading stage, keep searching using Agent Assist.
  3. After loading more than 200 plugins, verify the Agent Assist search.

Notice that the Agent Assist search is broken.

Agile Development


The Sprint Tracking tab/board in the Agile Board is not working

App Client UI


The plugin list continually refreshes if it is accessed by a domain separated user The plugin list continually refreshes if it is accessed by a domain separated user. Any filters selected are removed after the page refreshes.

App Client UI


The 'Update' button is not enabled for installation for the higher version



Email approvals are showing as desktop approvals Approvals that come in via emails are tagged as desktop approvals.

Appsec – Platform Dev


The introduction of Security Knowledge Base causes duplicate articles

Authentication - SSO


The certificate validation fails when the keyInfo field is missing in the certificate with Multi-SSO v2 After upgrading to New York, Multi-SSO v2 fails to authenticate users if the keyInfo field is missing in the certificate.

Authentication - SSO



SSO users are asked to 'Use External Login' if they exit the browser after logging into the instance The glide_sso_id cookie is removed upon browser closure. Therefore, SSO users are required to 'Use External Login' if they close the browser even if they have already logged into the instance once.
  1. Configure an instance with SSO, and do not set a default nor a default redirection to an SSO IdP.
  2. Log in to the instance for the first time.

    You will be required to select 'Use External Login'.

  3. Select 'Use External Login', then fill out the required information.

    You will be redirected to the IdP. Once you are authenticated to the IdP, you will be allowed to log in to the instance.

  4. Close the browser and reopen the browser, and try to log in to the instance again.

Notice that it will not automatically redirect you to the IdP. You will need to use 'Use External Login' again if you log in.

Automated Test Framework


ATF must properly wait until the webpage and all its internal frames are ready for the next test step UI steps could report an error when the page contains multiple frames and the inner frame executes asynchronous HTTP call and/or performs DOM mutation, for example, the rotating spinner appears in the inner frame during the page load.

Change Management


Cannot update or write to journal fields through Change Management REST API Users cannot update or write to journal fields through Change Management REST API with either POST or PATCH.

Cloud Management Application



The 'New' button is missing under the 'Cloud Templates' related list in the 'Cloud Catalog Item' form

Refer to the listed KB article for details.

Cloud Management Application


VMs are in the 'STOPPED' state instead of the 'Off' state
  1. Run the discovery.
  2. Navigate to the 'cmdb_ci_vm_instance' table.

Notice that some VMs are in the 'STOPPED' state.

Contextual Search



The filter configuration under the table configuration does not work after changing the searcher from the search context The contextual search is returning all knowledge results instead of specific results after the upgrade.

Refer to the listed KB article for details.

Core Platform


Issue with the time when creating a fiscal calendar after the New York upgrade
  1. Create a standard fiscal calendar.
  2. Navigate to the business_calendar_span table.

Notice that all the end times are 7 or 8 hours ahead of the expected end times.

Core Platform


The start and end input times of business calendar entries are converted from UTC to the user time zone
  1. Create a business calendar in the Pacific Time Zone.
  2. Create a calendar entry name for the calendar.
  3. Create a calendar entry for the calendar with the start time '2019-08-28 15:15:06' and the end time '2019-08-30 15:16:34'.
  4. Save.

Expected behavior: The start time should be 2019-08-28 15:15:06, and the end time should be 2019-08-30 15:16:34.

Actual behavior: The time is converted from UTC to the Pacific Time Zone. The start time is now 2019-08-28 08:15:06, and the end time is 2019-08-30 08:16:34.

Core Platform


In the business _calendar _span table, the display value on the list view of a non-floating span is showing the UTC value
  1. Create a business calendar in the Pacific Time Zone.
  2. Create a calendar entry name for the calendar.
  3. Create a calendar entry for the calendar with the start time '2019-08-28 15:15:06' and the end time '2019-08-30 15:16:34'.
  4. Save.
  5. Navigate to the business_calendar_span table and look for the span just created.

Expected behavior: The start time should be 2019-08-28 15:15:06, and the end time should be 2019-08-30 15:16:34.

Actual behavior: The time is converted from the Pacific Time Zone to UTC. The start time is now 2019-08-28 22:15:06, and the end time is 2019-08-30 22:16:34.

Core Platform


Issue with the enforcement of the entitled tables in the table drop-down for Mobile Studio data item

Customer Service Management



The 'New' UI action in the 'Companies' related list for 'Domain' table records is not visible in Madrid

Refer to the listed KB article for details.

Customer Service Management


Users who have installed OpenFrame but not the CTI Softphone plugin are not able to create interactions by using the scripted REST API endpoint on Agent Workspace

Customer Service Management


The OpenFrame 'init' event is timing out in the OpenFrame component in Agent Workspace

Customer Service Management


The sold product/install base in the record producer's drop-down is not shown when there is no many-to-many relationship with install base/sold product respectively

Customer Service Management


After upgrading to New York, the Virtual Agent case status fails to display any status of the case
  1. Make sure Virtual Agent is active with 'CSM Virtual Agent topics'.
  2. Click the chat icon.
  3. Click Show me everything.
  4. Check the case status.
  5. Search for 'my active cases'.
  6. Pick one active case.

Notice that there is no case information under 'Here's the case you selected:'.

Database Indexes


Snapshot tables do not have indexes on the primary key fields Snapshot tables do not have indexes on the primary key fields (for example, 'np$XXX.sys_id').




The error message 'Discovery log is too large and could not be saved' appears in the discovery log The error message 'Discovery log is too large and could not be saved' appears in the discovery log. Increasing the length of the sa_discovery _log.log dictionary does not solve the issue.

Refer to the listed KB article for details.



The password reset process generates log errors when trying to insert to the perf_metric_ probe_sensor table When the Password Reset plugin is enabled and no Discovery plugin is enabled, the log errors related to a non-existent table 'perf_metric_probe _sensor' appear after the password reset process.



This PRB fix addresses the issue where re-discovering a CI using patterns would assign a new IP to the CI’s ip_address field even when the old IP is still a valid IP for that CI



When running a discovery containing multiple launch parameters to the same pattern, the same launch parameter values are used for all of them
  1. Create two pattern launch parameters that launch the same pattern.
  2. Give two different values in the parameters.
  3. Run the discovery.

Notice that both launches use duplicate parameter values in the ecc_queue output payload rather than using the different values that were selected.



Probes with credential tags get empty affinities, which impacts the discovery performance Probes that use credential_tag are receiving credential_id as "" (empty) after the New York upgrade, which causes slowness in the discovery performance.

Domain Support


Strict dot-walking enforcement breaks some existing processes

Domain Support


Large table audits can be scheduled daily via list view Large table audits can be scheduled daily via list view by admin users. However, this should not be allowed.

Edge Encryption


When an Edge-encrypted record contains Windows special characters (for example, " ) or emoji, the Edge proxy fails to process this record when retrieving through REST API or Service Portal
  1. Install and configure the Edge Encryption proxy.
  2. Create the encryption configuration on incident::Description.
  3. Create a record containing the quote-like special character < " > (You can use Outlook mail to get the character).
  4. Try to retrieve this record through the REST API explorer or Service Portal.
Note: Viewing and editing the record in the HTML form is not affected.



KMF fails to pull certificates from vault

Event Management


The job 'Service Analytics group alerts using RCA' causes out of memory exceptions

Event Management


SNMP listener parser does not cut off trap counter from the OID name, and event rules cannot extract information from the additional info SNMP listener parser does not cut off trap counter from the OID name. This causes conflicts in the event rules to extract information from the event additional info.

Event Management


Em_events are stuck in the ready state because the cache is not refreshed correctly Em_events are stuck in the ready state because the cache is not refreshed correctly, but the number of jobs is correct.

Financial Management


Expense lines processing the account actuals amount in the incorrect fiscal period

Flow Designer


In Flow Designer, the 'Update Record' action uses the choice display value and not the choice value Flow Designer is using labels instead of values for the reference data pill picker. Therefore, when users try to update fields in a record that have numerical values (for example, the 'State' field), the values are not updated in Flow Designer.

Flow Designer


Using the caret character '^' in a value set with a Create Record action results in the set value being truncated at the caret character '^'
  1. Create a flow 'copy problem'.
  2. Set the trigger to problem created with the description 'copy me'.
  3. Add the Create Record action with the following information:
    • Table: Problem
    • Fields: Problem Statement = Trigger->Problem Record->Problem Statement
  4. Activate the flow.
  5. Create a problem with the problem statement 'testing^this^out' and the description 'copy me'.
  6. Verify that the flow runs and creates a problem with the problem statement 'testing'.

Flow Designer


Conditions in Flow Designer with unmet references are always evaluated as false When a condition contains a reference (pill) to an action that is not executed, the whole condition fails parsing and thus is always evaluated as false.

Flow Designer


The dynamic input value is not properly saved when running test in Flow Designer

Flow Designer


The + button on name value pairs is not visible in New York
  1. Navigate to Flow Designer.
  2. Create an action and add the REST step.

Notice that in the Query Parameters and the Headers sections, the + button is not visible. You need to click the step panel to make it visible.

Flow Designer


An error occurs when creating a flow to trigger on sysevent without any condition The following error occurs on the screen when creating a flow to trigger on sysevent without any condition: 'Error: Internal Server Error. java.lang. StackOverflowError'.

Flow Designer


Unable to retrieve data (Application, Collections, Deployments, etc.) using the Discover subflow
  1. Activate the following plugins:
    • Orchestration - Client Software Distribution
    • SCCM integration spoke
    • IntegrationHub Runtime
  2. Create the MID Server.
  3. Create the connection and the credential to the SCCM alias.
  4. Create a flow with the Discover SCCM subflow.
  5. Test the flow and verify the data.

Notice that the SCCM data are not populated in the respective tables.

Flow Designer


Dynamic inputs with the MID Server are failing

Flow Designer


Flows triggered with multiple RITMs on the same request (or as a part of the order guide) evaluate the trigger with the same sc_req_item record Flow Designer runtime values are showing incorrectly for RITMs.

Flow Designer


Unable to save the flow if the payload is too large

Flow Designer


The drag and drop does not work on Array.Object and object child string type
  1. Create an action with Array.Object/object outputs and string type input.
  2. Add the string type child to the above outputs.
  3. Try to drag and drop the input pill to the string type child.

Notice that the drag and drop does not work and the pill picker is missing.

Flow Designer


The browser hangs when viewing a flow in Opsview When viewing a flow in Opsview, the flow does not load and errors occur in the browser console.

Flow Designer


Unable to publish an action which has a specific complex object structure

Flow Designer


Values are not mapped to the nested object variables of a complex object which is created with a specific structure

Flow Designer


Values are not passed from a subflow input which is of the type Array.String
  1. Navigate to Flow Designer and create an action with the input of the type Array.String.
  2. Create a flow and add the subflow with the input type Array.String.
  3. Open the above created action, and drag and drop the input value from the subflow.
  4. Save the flow.

Notice that the subflow input which you drag and drop is removed whenever you save the flow.

Flow Designer


Action inputs object with duplicated inner child key reference issue

Flow Designer


Flows that are stuck for a long time are not automatically assigned to other free scheduler threads If there is a long running flow, it blocks all the other flows that are picked by the same event processor thread.

Forms and Fields


Submitting a form before a modified reference field has lost focus can cause data loss if there are dot-walked fields If users have a form that references another table and has dot-walked fields from it, data loss might occur in the dot-walked fields if users update the reference field and submit the form without clicking elsewhere first.

Forms and Fields



An empty dot-walked field on a form which is set to mandatory at the dictionary level does not allow users to save the record When a dot-walked field is added to the form layout and it is mandatory at the dictionary level, users cannot save the record if the dot-walked field is empty.

Refer to the listed KB article for details.

Forms and Fields


State choices are missing for task tables Choices are missing in the filter for the 'State' field in task tables.

Guided Tours


The guided tour in Service Portal requires users to authenticate even if they are already authenticated

Guided Tours


The 'Click' action event is not available to be selected for lists in the portal

Guided Tours


URLs (external links) are not working after the upgrade In a guided tour, the external link added in a step cannot be opened when clicked.

Human Resources Service Management


Changes are made for the modified APIs of SuccessFactors to push to-dos to SuccessFactors

Human Resources Service Management


Duplicate snc external criteria and roles on the installation of the HR LE plugin

Human Resources Service Management


Changes in the script include and the workflow cause upgrade issues

Human Resources Service Management


After upgrading to New York, existing flows related to the PDF generation break as images are expanding out of proportion After upgrading to New York, images populated in the PDF are expanded by more than 3 times as compared to Madrid.



The Microsoft AD spoke actions fail with errors The following error message appears in the log: 'Worker-Expedited: IPaaSActionProbe SEVERE *** ERROR *** Failed while executing ActionAdd UserToADGroup.ps1 (Access denied)'.



After the plugin 'Microsoft Teams Spoke for ServiceNow IntegrationHub' is enabled, the application picker does not have an option for 'Microsoft Teams'



Issues with the caller scope in transactions and feature usage The caller scope calculation needs to be updated for licensing and usage purposes.



The subflow fails when run from the script if it contains a Create or Update action



When the ProcessPlan JSON serialization fails, the error message in the log is often truncated or does not provide enough context



The connection alias that uses the retry policy fails to resolve the connection info if the MID user is missing admin rights



FlowExecutionWaiter is not notified in the case of multiple nodes where MID notifies the instance that the waiter is not on Synchronous flow or action execution will fail in multi-node environment when the flow or action executes on MID.



Incorrect indexing when picking the element of the complex object using the data pill



Unable to expand actions in the flow after adding the For Each flow logic

Knowledge Management


Setting array_denormalized to false can break functionality on the knowledge table

Knowledge Management


The header is missing when accessing Service Portal kb_home from the web page using 'target="_blank"'
  1. Create an HTML file that contains anchor links to open Service Portal:
    • <a href=" id=kb_home" target="_blank">Link to kb_home (_blank)
    • <a href="https:// id=kb_home"> Link to kb_home </a>
  2. Open the HTML file from the browser.
  3. Click Link to kb_home (_blank).

    Service Portal will be opened in a new window and there is no header.

  4. Click Link to kb_home.

Service Portal will be opened in the same window and the header is shown.

Knowledge Management


Default semaphore exhaustion due to the script include SocialQATags The script include SocialQATags creates a query against the label_entry table with no filters and causes out of memory.

Knowledge Management


If multiple language support is installed, users do not see all the articles they have access to With multiple language support, the knowledge widget has an option to show articles in other languages. This widget does not always show all the languages but only a subset of the ones users have access to. This is because the list stops when users click an article they do not have access to.

Knowledge Management


Article dot-walked fields on lists and reports are not displayed

Knowledge Management



The bottom of the page is displayed when opening a knowledge article When the system property 'glide. knowman.show_ user_feedback' is set to 'onClick', the bottom of the page is displayed by default when opening a knowledge article.
  1. Log in to a New York instance.
  2. Set the system property 'glide.knowman .show_user_feedback' to 'onClick'.
  3. Open a knowledge article that users have to scroll to view.

Expected behavior: The top of the page should be displayed by default.

Actual behavior: The bottom of the page is displayed by default.

Knowledge Management


The language filter selection in the knowledge portal is not applied when users refresh the page When users refresh the KB portal search page, the user language preference is not honored in KB search results.
  1. Activate any language plugin.
  2. Create/Edit a KB article and set the language to the one you activated (for testing purposes, include 'Email' in the title).
  3. Navigate to the /kb Service Portal and search for 'Email'.

    You should see a few results.

  4. Use the language filter to change the language to the one you activated.

    You should see the results refreshed and the KB you created in Step 2 should be shown.

  5. Click Home in the breadcrumb to go back to the homepage.
  6. Search for 'Email' again.

The results shown should still be filtered by the selected language shown on the left. However, if you refresh the page, you will get a set of results that are not filtered by the language, but the language filter to the left still shows one selected.

Knowledge Management


Issue with viewing knowledge articles in Service Portal Some ITIL users are unable to view KB articles in Service Portal, however, they are able to view the same article in the platform UI.

Language and Translations


Some of the system applications and the login page are displayed in Japanese when 'Traditional Chinese' is selected
  1. Log in to a London instance.
  2. Navigate to plugins and install 'Chinese' and 'Traditional Chinese' languages.
  3. Once the plugins are installed, upgrade the instance to New York.
  4. Open the login page.
  5. Change the language from 'English' to 'Traditional Chinese'.

    Notice that the username and password fields are displayed in Japanese.

  6. Log in to the instance.
Notice that some of the application modules are displayed in Japanese. If you navigate to the 'sys_ui_message' table, you will notice that the message for 'UserName' and 'Password' is in Japanese and the records were updated by the system when the instance was upgraded.

MID Server


MID Server threads hang on the execution of the 'Enable-PSRemoting-Force' command When multiple threads execute the 'Enable-PSRemoting-Force' command concurrently, one of the threads hangs indefinitely due to a lack of response from the PowerShell console.

Mobile - Service Catalog


Performance issues with 'Popular Services' in Mobile 'Popular Services' in Mobile triggers three slow queries when loading the homepage and other pages.

Mobile Studio (non-classic)


Saving existing applets in Studio results in applets having no data
  1. Activate the ITSM Mobile plugin.
  2. Open the ITSM applet in Studio.
  3. In the mobile app, navigate to Active Critical Incidents and Active Incidents applets.

    Notice that they have data currently.

  4. Navigate back to the browser and open the Active Incidents applet in Studio.
  5. Save the applet.
  6. In the mobile app, navigate to Active Critical Incidents and Active Incidents applets again.

Notice that now there are no data found.

Outbound Messaging


Errors occur for outbound REST transactions with payloads bigger than 864 KB The following error occurs for outbound REST transactions with payloads bigger than 864 KB when trying to process the response: 'Exception: Error while evaluating the XPATH Expression against response'.

Password Reset Application


Password reset flows are executed in the global domain for users with the guest session login

Password Reset Application


Emails with the verification code for the SMS device are set to SMTP notification _type and thus cannot be delivered to mobile devices

Performance Analytics Application


The element clicked from the related breakdown is not applied in indicator filter conditions in Analytics Hub

Performance Analytics


Encoding issue with the workbench widget

Performance Analytics


The Analytics Hub target progress bar does not refresh when clicking through scores
  1. Collect data for PA incidents.
  2. Open the 'Number of open incidents' indicator in Analytics Hub.
  3. Create a target in the past.
  4. Click the line chart through various scores.

Performance Analytics


The New York upgrade of the Performance Analytics (PA) content for Configuration Management (CMDB) may cause performance degradation



When deleting records from a table in a TPC hierarchy, a connection error can cause data loss If the schema of a table within a TPC hierarchy is rebuilt and there is a connection exception for the DB during the process, then later when records are deleted from a child of that table, it is possible for the platform to erase all records in the whole hierarchy.

Policy and Compliance


Users should not be allowed to reassign the child assessment Users should not be allowed to reassign the child assessment. Reassignment should only happen in the group level.

Project Management


Issues on the creation of the cost plan/benefit plan breakdown with zero amounts and later updated with a non-zero amount

Project Management


In New York, the planning console takes a long time to load when the overall planned duration of the project is large
  1. In a New York instance, create a project record.
  2. Add a project task with a planned start date and planned end date in January 2016.
  3. Add a project task with a planned start date and planned end date in December 2019.

    The planned duration for the project should be over 1000 days.

  4. Load the planning console.

Notice that it takes 20-30 seconds to load even though there are only 2 project tasks and no dependencies.

Record Watcher


The record watcher responder does not work even though the GlideRecord condition is met

Resource Management


The weekend time card hours are not captured in the resource plan and resource allocation actuals

Resource Management


Floating point or round off issues in the resource form, resource workbench (or allocation workbench), and even allocation engine

Service Catalog


Incorrect Requested for value in Service Portal
  1. Log in as an admin user.
  2. Navigate to Service Portal.
  3. Add an item to the cart.
  4. Navigate to the cart in the header menu and click View Cart.
  5. Click Proceed to checkout.
  6. Pick a user for the Requested for field, for example, Abel Tuter.
  7. Click Cancel.
  8. Navigate back to the catalog and open any item with the request method 'Submit'.
  9. Click Submit.

Notice that the request is requested for the user you picked in Step 6 and not yourself.

Service Catalog



Ordering a catalog item fails with an error when twostep is set to false When is set to false, ordering a catalog item fails with the error 'FAILURE: Validations failed. Failed to Order Item'.
  1. Create an ATF test with two steps:
    1. Open a catalog item (SP) with the Apple iPad 3 catalog item.
    2. Order a catalog item (SP) and assert that it should succeed.
  2. Run the ATF test.

    It should pass.

  3. Change the value of the system property '' to false.
  4. Run the ATF test.

In the second step of the test, it fails with the message 'FAILURE: Validations failed. Failed to Order Item.'

Service Catalog


Setting a 'null' value to a variable breaks the 'Submit' function of the record producer If the value for a question is set to 'null' using a catalog client script and if this variable/question is used in a record producer, the user cannot submit the record producer.

Service Catalog



The Requested Item record cannot be edited if the multi-row variable set in the variable editor is set to Read-Only by a catalog UI policy
  1. Create a catalog item 'TestSNC'.
  2. Create a single-line text variable 'Q1'.
  3. Create a multi-row variable set 'MRSV'.
  4. Create two single-line text variables 'Name' and 'Age' for 'MRSV'.
  5. Create a catalog UI policy 'Test UI Policy':
    • Enable 'Applied on Requested Items', 'On load', 'Reverse if false' and 'Run Script'
    • Execute if true: 'g_form.setReadOnly('mrsv',true);'
  6. Request the catalog item 'TestSNC' on Service Portal.
  7. Click to open the requested item in the platform view.
  8. Add additional comments, change watch list, state quantity, Q1.
  9. Click Save.

Expected behavior: Fields in Step 8 should be saved on the form.

Actual behavior: Fields in Step 8 are not saved on the form.

Service Level Agreement (SLA)


If an SLA async processing job has to update a record with the state of 'error', it will not process any other jobs allocated to it and leave them 'queued' If an SLA async processing job has to update a record with the state of 'error', it will not process any other jobs allocated to it and leave them 'queued'. This can lead to a build-up of records with the state of 'queued' in 'sla_ async_queue', and they will never be processed and can eventually result in no SLA processing.
  1. Enable SLA async processing by navigating to the module 'SLA Engine' and checking the property for 'Execute the 2011 SLA engine asynchronously'.
  2. Navigate to the table 'sys_trigger' and find the record with the name 'SLA Async Delegator'.
  3. Update the type of this record from 'Repeat' to 'On demand'.

    This will temporarily disable the allocation of SLA Async Queue records to background jobs.

  4. Create 5 Incident records.
  5. Navigate to the table 'sla_async_queue' and confirm there are 5 records for the 5 incidents you just created.
  6. Update the record with the lowest 'Sequence' number to have an empty 'Update number' field.
  7. Navigate back to the table 'sys_trigger' and find the record 'SLA Async Delegator'.
  8. Update the type of this record from 'On demand' back to 'Repeat'.
  9. Navigate back to the 'sla_async_queue' table.

After a few seconds, you will see that the record with the lowest 'Sequence' number has been updated to have the state of 'error' and the other 4 records have the state of 'queued', but the 'SLA Async Job' field is empty.

Service Level Agreement (SLA)


Async SLA processing can be delayed or stopped completely due to one or more records with a state of 'processing' in the 'sla_async _queue' table but with an empty 'SLA Async Job' field

If SLA Async processing processes a record out of 'sla_async_queue' and the node responsible for running the job to perform the processing is restarted, this record will be left in a state of 'processing' but with an empty 'SLA Async Job' field. This means it will never be processed or updated by SLA processing and also reduces the number of jobs available to process 'sla_async_queue' by one job.

This can result in delays in SLA processing, and if this happens 4 times, no SLA processing will be performed (SLA async processing is limited to a maximum of 4 concurrent jobs).

Service Mapping


The sm_admin role needs to have access to the cmdb_ key_value table
  1. Log in with the role sm_admin.
  2. Navigate to cmdb_key_value.list.

Notice that the sm_admin role has no access to cmdb_key_value.

Service Owner Workspace


Updating the state of the Service Owner Workspace plugin

Service Portal


When adding an image to the HTML input for the second time, canceling the file upload window hangs the editor In Service Portal, after adding an image to the HTML input and starting to add a second, canceling the file upload window hangs the editor.

Service Portal


The reference field auto-complete results should be cached while on the same form In Service Portal, the reference field auto-complete results should be cached while on the same form so that the same query is not made to the server multiple times.

Service Portal


The UI action interaction on the Form widget does not extend the user session timeout
  1. Log in as the admin.
  2. Create a table 'Test' with the name 'u_test'.
  3. Add the following two fields:
    • Name, String, maxlength 20
    • Short description, String, maxlength 250
  4. Create a u_test record.
  5. Set the system property 'glide.ui. session_timeout' to 2 (the value is in minutes).
  6. Log out and log in again.
  7. Open the u_test list in Service Portal.
  8. Click to view the u_test record created in Step 4.
  9. Every 60 seconds for the next 5 minutes, modify the field Short description and click Save.
  10. Wait for another 1 minute.

Expected behavior: The user session should still be active after 5 minutes.

Actual behavior: The user session times out at the second minute because updating the record does not reset the session timeout.

Service Portal


In New York, the search page does not work properly in IE11 The Service Portal search page does not work properly in IE11 after the New York upgrade.
  1. In IE11, log in to a New York instance and navigate to the search page on Service Portal.
  2. Search for a term.
  3. Change the search term and press enter.

Expected behavior: The search results should be updated based on the new search term.

Actual behavior: The search results are not updated based on the new search term.

Service Portfolio Management


Read/write roles are not being inherited when accessing the service _offering tables after the New York upgrade

Software Asset Management Professional


The sam_admin /sam_user cannot access Software Model Lifecycle Report
  1. Create a user with the role sam_admin /sam_user.
  2. Navigate to Reports > View/Run.
  3. Search for the report 'Software Model Lifecycle Report'.

Notice that the sam_admin/sam_user is not able to view this report.

Software Asset Management Professional


The product documentation link needs to be updated on the Content Service Setup page The current product documentation link on the Content Service Setup page is the home page of the product documentation website. It should link to the Spend Detection Content Service page.

Software Asset Management Professional


Previously imported staging rows are parented to the incorrect upload, causing unintentional deletion of data
  1. Enable the Software Spend Detection plugin.
  2. Import Spreadsheet A into Software Spend Detection.
  3. Import Spreadsheet B into Software Spend Detection.
  4. After all imports/predictions have completed, delete the import results record from Spreadsheet B.

Expected behavior: The staging rows for Import A are still in imp_sam_spend_transaction.

Actual behavior: The staging rows for Import A are deleted when deleting Import B.

Software Asset Management Professional


Running reconciliation results in unlicensed installs if there are two records for the same user and one is active while the other one is inactive The subscription pull job normalizes user principal name with inactive users, but reconciliation does not. Running reconciliation may create unlicensed installs if there are two records for the same user and one is active while the other one is inactive.

Software Asset Management Professional


Upgrading to New York takes a long time because of the fix script that creates tasks for transfer _order_line records For instances that have transfer _order_line records, upgrading to New York takes a long time because the fix script needs to complete creating tasks for transfer_order_line records before the upgrade can be completed.

Software Asset Management Professional


The script is likely to run in excessively long duration and hence the system would suspend its execution

Source Control Integration


Switching branches can cause sys_metadata _deletes to be incorrectly created for choice lists
  1. Create an application and commit it to source control.
  2. Add a table with a choice field that has at least one choice.
  3. Commit the new table to source control.
  4. Create a branch in source control.
  5. Verify that no sys_metadata_delete records exist for your application.
  6. Switch back to master.
  7. Verify that a new sys_metadata_delete record is created for your choice list, and verify that the choice record is still present for your field in the related list of the dictionary entry for your field.
  8. Add a new file to your app, for example, a business rule or a script include.
  9. Commit all of your changes.
  10. Verify that your app's author_elective_update folder now contains the sys_metadata_delete for your choice field.
  11. Switch the branch to the branch created in Step 4.
  12. Switch the branch back to master.

Notice that your choice field no longer has any choices.

Vendor Manager Workspace


Activation issue with the Vendor Manager Workspace plugin

Virtual Agent Platform


The Virtual Agent (VA) keyword search for topics with stop words throws an error and the topic search does not work If a topic has a keyword that contains a stop word, the following error appears on all tables in the platform during the search: 'Your text query contained only common words or ambiguous wildcards, please refine your search and try again'.

Virtual Agent Platform


The CSM Workspace plugin is activating the Predictive Intelligence plugin

Visual Task Boards


The 'Lane limit Exceeded' notification is displayed when not required The notification 'Lane limit Exceeded' appears in Visual Task Boards when no lane limits are set.
  1. As the admin, create a freeform board.
  2. Add the ITIL user, and add a new card to the To Do lane.

Expected behavior: No VTB notifications should be triggered.

Actual behavior: The VTB notification 'Lane limit Exceeded' is triggered.

Visual Task Boards


Unable to set the primary assignee on the guided VTB for the task table with sys_class_name as the field When users try to set the assignee of a card in a guided board on the task table using the Task Type field for lanes, an error message appears: 'An unknown error has occurred'. However, it seems that the 'Assigned to' field is still updated.

Walk-Up Experience


When scheduling an appointment, the interaction record which is supposed to be 'On Hold' sometimes changes to 'New' In the walk-up portal, when scheduling an appointment, the state of the interaction record changes from 'On Hold' to 'New' intermittently.
  1. Log in as a walk-up user and navigate to the walk-up portal.
  2. Choose a location.
  3. Click Schedule appointment.
  4. Pick an available time from the picker and click OK, then click Schedule appointment.
  5. Navigate to the list of active interactions and check the state of the newly created interaction.

Notice that it is set to 'On Hold'. After the scheduled job 'Interaction state change for walk-up' gets executed, the State field of that interaction record changes to 'New' even though the appointment is in a future date/time.

Note: A few attempts might be needed to reproduce the issue.

Web Services


Duplicate login windows appear for IE11 In IE11, after 'Session Timeout', duplicate login windows appear on the screen. Users can still use the top window to log in.



Deleting a development application (sys_app) deletes wf_workflow _instance records required by other applications

When a workflow contains another workflow, a wf_workflow_instance record is used to reference the sub-workflow. If the referenced sub-workflow is from a development scoped application, which means it is a sys_app and not a sys_store_app, any wf_workflow_instance records referencing that workflow will be deleted when the app is deleted.

These records do not come back when it is reinstalled, which means the other workflows referencing the deleted one will be permanently broken unless the wf_workflow_instance records are manually restored.

This only applies when deleting sys_app applications. If a sys_store_app is uninstalled, it does not delete wf_workflow _instance records and this issue does not occur.



The Wait for Condition step/action does not resume the flow even though the GlideRecord condition that it is waiting upon is met

Fixes included

Unless any exceptions are noted, you can safely upgrade to this release version from any of the versions listed below. These prior versions contain PRB fixes that are also included with this release. Be sure to upgrade to the latest listed patch that includes all of the PRB fixes you are interested in.