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

New York Patch 1

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

New York Patch 1

The New York Patch 1 release (General Availability) contains important problem fixes.

New York Patch 1 was released on September 4, 2019.
Build date: 08-24-2019_0944
Build tag: glide-newyork-06-26-2019__patch1-08-14-2019
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 1 includes 135 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 1 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 1, refer to KB0759084.

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

Service Catalog: Service Portal Widgets



The 'Required information' section is empty when the title of the variable set is the same as the name of a mandatory variable An error appears in the console panel, and the 'Required information' section is empty.

Refer to the listed KB article for details.



Thin columns appear in column, trend, and bar reports, which makes them difficult to use and prevents the ability to drill into the data

Mobile (Classic)


Incorrect timestamps for activity streams show up in the ServiceNow Agent mobile app when the timezone is set to various values, for example, Europe/London, US/Eastern, and Europe/Amsterdam
  1. Change the following system property to the value 'Europe/London'.
  2. Clear cache.
  3. Assign an incident or wm_task to the system administrator, and set the timezone field value of the system admin sys_user record to 'System' or 'Europe/London'.
  4. Log in to the ServiceNow Agent mobile app as the above system administrator, and access the above record.
  5. Add a work note.

Notice that the timestamp registers as ''Europe/London' + 3 hours'. The ServiceNow Agent mobile app shows incorrect timestamps ('Europe/London' + 3hours) for activity streams when is set to 'Europe/London'. However, if you open the same record from desktop, the timestamp correctly shows as 'Europe/London', which is the user's current timezone in the instance.

Asynchronous Message Bus



The Asynchronous Message Bus (AMB) JS client does not correctly re-subscribe to channels after a disconnect If users navigate away from Safari to another app such as Messages or Email and then back to Safari after a minute, the AMB will not be correctly connected and new messages will not show.
  1. Open two windows with different users and start a chat.
  2. Disconnect the internet for more than 30 seconds.
  3. Reconnect the internet.
  4. Try resuming the chat between the two users.

Notice that the second user does not get the message you attempt to send.

Core Platform


Attaching files via the attachment API results in the file descriptor leak leading to the node crash Attaching files via the attachment API results in the file descriptor leak and 'Too many open files' exceptions.



Populating the Assigned to field on the Incident form is slow

All other fixes

Problem Short description Description Steps to reproduce

Agent Intelligence


The 'Output Field' field of the classification solution definition is not immutable when training the solution

Agent Intelligence


Clustering prediction failure: The trained model does not have active associated word vector corpus
  1. Navigate to Predictive Intelligence > Word corpus.
  2. Create a word vector corpus with the name 'wc'.
  3. Navigate to Predictive Intelligence > Similarity > Solution definitions.
  4. Open ml_sn_global_similar_incidents.
  5. Select the 'wc' word corpus and submit the training.
  6. When the training is in progress, navigate to Predictive Intelligence > Clustering > Solution definitions.
  7. Create a solution definition by selecting the same 'wc' word corpus.
  8. Submit the training.
  9. Wait until both the similarity and clustering solution trainings are completed. Notice that the clustering solution's associated word corpus vector is empty.
  10. Try the clustering solution prediction from the REST API explorer.

Notice that all the predictions fail with the 'Unable to predict' message.

Agent Intelligence


The copy solution is not working
  1. Create a classification/similarity/clustering solution.
  2. Right click and select copy solution.

Everything except the label should be copied, but it's not working as expected.

Agent Intelligence


The similarity solution import from the update set is failing

  1. Get a trained similarity solution. Export to .xml using the update set.
  2. In another instance, import the solution in the update set.

When trying to import the similarity solution and running the preview, issues occur with word_vector.

Agent Intelligence



The machine-learning prediction 'mlpredictor' does not point to the correct URL

Refer to the listed KB article for details.

Agent Intelligence


Applying the similarity filter skips valid top results

Agent Workspace


After clicking Close Other Tabs, the workspace breaks
  1. Log in to the workspace.
  2. Click + to open two different interactions.
  3. Save the first interaction, and open the assigned_to preview record.
  4. Save the second interaction, and open the assigned_to preview record.
  5. Open two more tabs.
  6. Navigate to one of the interactions and make some change.
  7. Navigate to another tab and click Close Other Tabs.
  8. Click Close.

Expected behavior: Only one tab should be displayed.

Actual behavior: The workspace UI breaks and 'Loading...' is displayed on the page.

Agent Workspace


The choice field is resetting to the default value

Agent Workspace


All tabs are not closed after clicking 'Close All Tabs'
  1. Log in to workspace.
  2. Open any incident.
  3. Open a subtab for the incident.
  4. Make some changes to the subtab.
  5. Click Close All Tabs from the menu option.

    A confirmation message should be displayed.

Expected behavior: All tabs should be closed.

Actual behavior: The tabs does not close.

Agent Workspace


Workspace font 'SourceSansPro' is not loaded

Agent Workspace


Workspace doesn't load and generates console errors

App Client UI



The sys_app form is missing annotations for Application Scoping - Design - Runtime information on the record fields The sys_app form is missing annotations for Application Scoping, Design and Runtime, etc. that provide users with insight into the purpose of the record fields.

Refer to the listed KB article for details.

Application Portfolio Management (APM)


Unable to add the hint to the business application fields when the field label is against the base cmdb_ci table



Cannot pass a dynamic 'list' of users for approval into the 'Ask for Approval' action While it is possible to select a 'field' of the type 'list' as a dynamic feeder for the 'When' clause of a rule in the 'Ask for Approval' action, it is not possible to pass a parameter of the type 'list' there, which means the approvers, if multiple, cannot be dynamically built in a flow.
  1. Open Flow Designer.
  2. Navigate to the Actions tab.
  3. Click New > Create New Action.
  4. Create a custom action which contains the sys_id of users that are going to be considered as approvers or the logic to dynamically generate the list of approvers.
  5. Create a flow in Flow Designer and set the trigger to 'Service Catalog'.
  6. Use the action created in Step 4.
  7. Add the 'Ask for Approval' action.
  8. Click the 'Ask for Approval' action, and try to set the list from Step 6 in 'Anyone Approves'.

Notice that the error 'List is not allowed here.' occurs, and you are not able to pass the list generated from an action as an input.

AWA: Routing and Assignment


The work item state is not updated when the interaction is manually unassigned and in the new state
  1. Activate and load the demo data for the Walk-up Experience plugin (com.snc.walkup).
  2. Navigate to Advanced Work Assignment > Presence States.
  3. Open the Available state.
  4. Add a walk-up to the service channel and enable the active flag.
  5. Create a sys_user with the sn_walkup.walkup_technician role.
  6. Add this sys_user to one of the demo data assignment groups (either San Diego Tech Lounge or Santa Clara Tech Lounge).
  7. Impersonate the walk-up user.
  8. Navigate to Walk-up Experience > Walk-up Portal.
  9. Select a location.
  10. Click Check-in and fill out the form.
  11. Impersonate the user with the walkup_technician role and open Agent Workspace.
  12. Set the status to 'Available'.

    The work item for the interaction should appear in the inbox.

  13. Accept the interaction.
  14. Click the ellipses icon and click the Re-queue UI action.
  15. In another window, navigate to the awa_work_item list.

    The work item for the interaction should have a state of 'Accepted'. After five minutes, the interaction state should be reset from 'On Hold' to 'New'.

Notice that a new work item for the interaction is not created and thus does not appear in the agent's inbox.

Change Management


Approvals are not generated on Change Management when the language of the user is a language other than English Change approval policies do not work if the user sets the language to any language other than English.

Change Management


The change approval policy for the field type 'List' is not working when more than one user is selected
  1. Create a field of the type 'List' on the table cmdb_ci.
  2. Make sure that a CI has multiple users selected for this new field.
  3. For the 'Assess Technical Approvals' decision, make sure for the answer, 'approval action' is 'add a user approval'.
  4. For the user field, dot-walk from the selected CI on the change_request to this new field of the type 'List'.
  5. Submit the change request with the CI with multiple users selected on the new field.

When the workflow gets to the change approval policy, only the first user of the field is picked up. However, all users in the field should be picked up for approvals.




Random URL links show in the workspace chat window When creating or handling a case through the live agent chat, extra random URLs are generated in the requestor workspace chat window.

Refer to the listed KB article for details.



Unable to open an interaction when another interaction is open
  1. Log in to the workspace.
  2. Click + to open a new interaction and save it.

    The interaction record should be displayed.

  3. Click + again and try to click any option in the drop-down list.

Expected behavior: A new tab should be opened.

Actual behavior: The option in the drop-down list is not clickable.

Cloud Insights


The Insights installation should not disable CMP AWS billing The installation of Cloud Spend AWS automatically disables the CMP billing schedule, which is not expected.

Cloud Insights


Remove the cyclic dependency associated with the object instantiation to make the sysID cache accessible to corresponding instances

Cloud Management Application


Semaphore exhausted and high instance response time due to the frequent execution of the job 'Fill in tag assignments'

Cloud Management Application



Inaccurate reports cause the cloud billing to not set the cost field properly The following two issues have been observed:
  • When the Azure instance_id ends with '/', some costs are not assigned to the owner.
  • When the locale is different from the system default, the cost and usage quantity fields are incorrectly filled.

Reproduce Issue 1

  1. Create a schedule that pulls azure billing data.
  2. Find some instance_ids that are ending "/".
  3. Modify the query accordingly to validate the data.

Reproduce Issue 2

  1. Change the system local to
  2. Pull either AWS or azure billing data.
  3. Compare the cost with the cost which was downloaded by the system locale.

Core Platform


The data stream action generates errors at a splitter step before reaching the last page

Customer Service Management


Performance issue with event processing: PCR propagation to child cases for MOT use cases

Customer Service Management


UI Policies added in Proactive Operations plugin which hides a few fields (state, alert, description) in Case view

Delegated Development


Delegated developers can gain access to create update sets and use the update set picker
  1. Open Studio.
  2. Create a custom application.
  3. Click File > Manage Developers.
  4. Search for Abel Tuter and add as a developer.
  5. Assign the 'Reporting' permission.
  6. Log out and log back in as Abel Tuter.
  7. Open the settings from the upper right corner.
  8. Navigate to the Developer tab.

    Notice that the developer can access the Update Set Picker button and enable the update set picker. However, users should not be able to see the Update Set Picker options.

  9. Navigate to sys_update_set.list.

    Notice that the developer can see the New button on the table and can click to add new update sets. However, users should not see the New button.

  10. Navigate back to Studio and select the application.
  11. Open the Manage Developers window.
  12. Assign the All Metadata permission to the developer.
  13. Log out and log back in as the developer.

Notice that the developer has access to read the records in sys_update_set. However, the developer should not be able to see any records in sys_update_set. The developer can also see the update sets in the update set picker, but the developer should not be able to set the update set picker or the update sets. Only a delegated developer with the Manage Update Set or Publish To Update Set permissions should have access to the update set table.

Dev At Scale


Applying the stash changes option is not loading files into author_elective



Solaris server pattern is not populating relationships and not triggering ADM for zones




The 'SNMP - Routing' probe launched for BGP routers causes out of memory errors on the MID Server

Run discovery on any BGP router and see that 'SNMP - routing' is being triggered.



Help the Help Desk does not classify computer as 'Windows Server' when the Discovery plugin is activated When the Discovery plugin is activated, the Help the Help Desk script reports the CI as 'Computer' rather than 'Windows Server'.




The discovery of member accounts fails when the master account is configured to use IAM Instance Profiles When a master-member account hierarchy is present in AWS and uses IAM Instance Profiles on an MID Server running in the master account, the discovery fails to run in the member accounts even if the AssumeRole functionality is properly configured to allow master-member traversal.
  1. Set up a master-member hierarchy with AssumeRole relationship in AWS.
  2. Set up a MID Server on an EC2 instance in the master account.

    Set up an IAM role with required policies in the master account.

  3. Assign this new role from #3 as the IAM instance profile to the EC2 instance.
  4. Discover the master-member hierarchy from the instance.
  5. Attempt to discover resources within the member accounts.




The error 'ReferenceError: "gs" is not defined.' appears after running the discovery After upgrading to New York, the error 'ReferenceError: "gs" is not defined.' appears in the logs when running the discovery.

Refer to the listed KB article for details.

Edge Encryption


The cache of the keys is cleared on the instance causing the proxy to pull the keys and re-register itself The cache of the keys is cleared on the instance, causing the proxy to pull the keys and re-register itself to perform the validation of keys among proxies. This causes the Edge proxy to re-register every time the cache is cleared.

Edge Encryption


When using equality preserving on versions older than Madrid, upgrading to Madrid breaks the existing equality preserving in the table



The legacy notification preferences page does not load When ACLs are present which restrict the visibility of a notification, the legacy notification preferences page does not load. However, the current UI16 version of the preferences does load and works.
  1. Make sure that glide.notification.preference.ui.enabled is set to 'false'.
  2. Create an ACL that prevents the visibility to a particular subset of the notifications.
  3. Impersonate a non-admin user.
  4. Navigate to the user profile and click the notification preferences.

Notice that the page does not load, and there are errors in the console log.

Event Management


Flickering appears in alerts when opening in WS When users open any alert in WS, there is flickering in the ribbon on the left side. The data (number of alerts) in the ribbon on the right side (timeline) are not correct sometimes.

Event Management


Reduce the number of alerts from 10,000 for the grouping job and add protection for graph reuse

Event Management


Impact Calculation: Changing the contribution percentage for the application impact rule affecting service does not work if the service severity is 'Info' before the change When changing the percentage of the application impact rule to 50 and after that to any number above 50 but under 100 (the sum should be more than 100%), the business service will become the severity of the CI that the impact rule is changed on.

Event Management


The Event Management dashboard is not loading and multiple Event Management impact jobs are stuck The following issues have been experienced:
  1. The dashboard and/or the alert panel are not loading until the timeout exception.
  2. The next Event Management jobs might be stuck (running for a few hours):
    • Event Management - Impact for service groups
    • Event Management - Backfill Impact Tables
    • Event Management - Impact Calculator for BS

Event Management


On the Event Management dashboard, no alerts are viewed in the alert panel when there are service groups in the system The Event Management dashboard should show all of the alerts associated with the services. However, the alerts are not shown if there are service groups other than 'All' in the CMDB.
  1. Create a service group.
  2. Send an alert on one of the services.
  3. Open the dashboard.

Notice that no alert is shown in the panel at the bottom.



Issues with the related list conditions when running a report The following issues with the related list conditions occur when running a report:
  • When using multiple 'contains' conditions with an 'OR' condition, only the first condition is considered.
  • When using an 'is' operator with an 'OR' condition and dot-walking, only the first condition is considered.

Same issues are observed from a list view with List v3 enabled.

Flow Designer


The search functionality on the Flow Designer form is not working for the Get Attachments on Record action
  1. Log in as the admin.
  2. Change the domain TOP/ACME.
  3. Navigate to Flow Designer.
  4. Create a flow or add an action to any existing flow.
  5. Click Action and search for 'Get Attachments on Record'.

Notice that the search does not bring any results, however the record is there.

Flow Designer


Editing nested subflows results in incomplete reporting When chaining nested subflows, editing a subflow causes the flow report to show the nested subflow as not run.
  1. Create and publish a subflow S2 that logs 'hello'.
  2. Create and publish a subflow S1 that logs 'hello' and then calls S2.
  3. Create and activate a flow F1 that calls S1.
  4. Use the Test button to run F1.
  5. Edit S1 so that it logs 'hello2'.
  6. Edit S2 so that it logs 'hello2'.
  7. Use the Test button to run F1.

Notice that the second report shows the log action of S2 as not run.

Flow Designer


Type label issue when saving action inputs
  1. Create an action with the input structure: Object - String.
  2. Publish the action.
  3. Change the string input to choice, and then publish.

Notice that it still comes back as string.

Flow Designer


The Complex Object_Output step is not displayed for a specific complex object structure

Flow Designer


The output choice list for the Ask For Approval action disappears when the flow is published Within the Do the following until flow logic condition builder, if the until condition is configured to use the output of the Ask For Approval action in the DoUntil loop, the user is unable to load the Ops view when the flow is in the waiting state. The page is hanging and the console shows an error. Also, after the flow is activated, the drop-down list of the Approval State is empty.
  1. Navigate to Flow Designer and create a flow.
  2. Add the Incident Created trigger.
  3. Add the Do the following until flow logic and click Done.
  4. Add the Ask For Approval action to the DoUntil loop with the following settings:
    • Record: Trigger->Incident Record
    • Approve when Anyone approves: Trigger->Incident Record->Caller
  5. Click the until link and configure:
    • Condition 1: [1.1->Approval State] is Approved (Notice that the drop-down list is available when the flow is created.)
  6. Save and activate the flow.
  7. After activating the flow, click the 'until [1.1->Approval State] is approved' link to expend the condition builder.

    Notice that the Approval drop-down list is not available anymore.

  8. Insert the Incident record to trigger the flow.
  9. In Flow Designer, click the Executions button on the Flow page, and open the execution view that is in the waiting state.

Notice that the page shows the 'Loading' message but never loads the page. The console shows the error 'Cannot read property 'value' of undefined'. After the user approves the Approval request, the flow completes successfully.

Flow Designer


An error occurs when publishing an action with the action output having nested objects
  1. Create an action with any step, and define an action output, for example, 'Users' of the type Object.
  2. Inside the 'Users' object, create an attribute 'Links' of the type Object.
  3. Assign the 'Users' output to any of the steps output.
  4. Publish the action.

Notice that an error occurs saying 'Action Output: invalid variable name or missing reference or missing default value'.

Flow Designer


The input label automatically changes to small letters when the input has the type as Dynamic Choice
  1. In Flow Designer, create an action in any scoped app.
  2. Create an input with the type Dynamic Choice and with a name that has capital letters and spaces, for example, TEST NAME or Test Name.
  3. Save and publish the action.

Notice that the label changes to small letters with underscore ('test_name').

Note: This is observed only when the input has the type as Dynamic Choice and is the first in the order.

Flow Designer


The 'ProcessAutomationException' error occurs when trying to invoke the Zoom actions programmatically using the FlowAPI

Flow Designer


Upon activating/publishing a flow or subflow, the tree structure in the data pill output of the Look Up Record action goes away
  1. Create a flow or subflow.
  2. Add the Look Up Record action with the Incident table. You should be able to see all the columns inside the Incident table in the tree structure.
  3. Activate/Publish the flow or subflow.

Notice that the tree structure disappears, and the table name (incident) is replaced with record.

Flow Designer


Errors occur when saving Data Stream actions with the complex object output having more than a certain number of attributes After creating a Data Stream action with the complex object output, when a certain number of attributes are added to the complex object, saving the Data Stream action errors out with the following messages: 'Failed to deserialize ComplexObject Unexpected end-of-input in field name', '[truncated 7500 chars]'.

Flow Designer


Undesired fields appear in Array.Object

When an array of objects is passed into an output variable of the type Array.Object, this Array.Object adds extra data to it.

  1. Create a REST service which returns JSON like [{"name":"John","phone":"555"},{"name":"Peter","phone":"999"}].
  2. Create an action with a REST step that consumes the above service.
  3. After the REST step, add a script step to parse the responseBody to an output variable of the type Array.Object.
  4. Add an action output variable of the type Array.Object and drop the script step's pill into it. Test the action.
  5. Check the execution detail.

Flow Designer


CO is converted to GlideType string when the media type is not json or xml CO is converted to GlideType string when the media type is not json or xml. It can be */* which accepts any media type. In the case of */*, it should default to the media type json so that the CO gets created and not defaulted to the type string.

Flow Designer


A warning message appears when selecting an action with a dynamic input
  1. Open Flow Designer.
  2. Select the Create Record action from the Salesforce spoke.
  3. Add the action in the flow.

Notice that a warning message appears at the bottom right corner, which disappears after a minute. This happens to any action that has a dynamic input.

Flow Designer


Flow Designer doesn't work as expected

Forms and Fields


Adding a choice with a space results in a visible '%20' in the choice
  1. Open an Incident record (or any other form with a choice field).
  2. Right-click Contact Type and select Configure Choices.
  3. Add a choice, for example, 'Has Space'.
  4. Click Save.
  5. View the choices on the form or the sys_choice entries.

Expected behavior: The choice value and label should have a space in them (e.g., 'Has Space').

Actual behavior: The choice value and label have a '%20' in place of the space (e.g., 'Has%20Space').

Graph API


GraphQL $$parent$$ notation to access outer parent objects is not thread-safe

Currently in GraphQL, references to parent objects can be shared amongst more than one thread when multiple GraphQL queries using the $$parent$$ notation are sent in parallel. They share the same parent context when they should each have their own.

If one (usually large) request containing the outer parent object reference is made and another is kicked off while the first one is still running, the second one shares the same parent object reference as the first. This means that when the first one completes and it calls clearParent(), thread two's reference to this object is gone and an exception is thrown.

Human Resources Service Management


Three Edge rules do not have valid signatures

Instance Data Replication feature


Instance Data Replication is not handling failover correctly due to the incorrect batch size when checking for the last message

Instance Data Replication feature


IDR Plugin - com.glide.idr- has a state of 'Development' that is prohibiting Support from activating it for New York



If use_flow is enabled on the store type, the Credential Store form should have the hostname non-mandatory and the reference to connection alias
  1. Make sure Use Flow is enabled.
  2. Install Orchestartion add-on, which will install the AD and Integrationhub Runtime.
  3. Navigate to Password Reset > Credential Stores.
  4. Try to create a Credential Store.
  5. After selecting the Credential Store Type for which Use Flow is enabled, "Host Name" should be modified to non-mandatory field.



Remove the 'Deprecated' tag from the IntegrationHub runtime plugin name

Language and Translations


Added translation updates for 22 languages See Activate a language for more information.



Float values converted to double on clotho REST get/transform The double floating point error occurs in JSON REST response on MetricBase REST transform calls.

MID Server


The wrapper.jar version is old (3.5.17) for the glide-dist-upgrade wrapper after upgrading Tanuki to 3.5.36

MID Server


The MID installed as a daemon with systemd is unable to start after the update
  1. Download the MID to Linux OS that has systemd (check for /etc/systemd).
  2. Install the MID with sh agent/bin/ install.
  3. Start the MID with sh agent/bin/ start (just sh also works).

    Notice that there should be a mid.service in /etc/systemd/system.

  4. Upgrade the MID via:
    1. pinning the MID to another version and clicking Update MID in the related links.
    2. pointing the MID to another instance and restarting the MID.

Expected behavior: The MID should upgrade and start as normal.

Actual behavior: The MID does not start on its own after upgrading.

Mobile Backend (non-classic)


Some records are skipped when paginating

Mobile Backend (non-classic)


Hidden master detail screens are transformed into visible list screens The Madrid migration transforms master detail screens into a list screen and a form screen. If the master detail screen is marked as hidden, the resultant list screen should not be visible in the corresponding icon section, but currently it is.
  1. In Madrid, create a custom master detail screen which is hidden.
  2. Upgrade the instance to New York.
  3. Open Mobile Studio.
  4. Find the master detail screen in the list of applets and open it.
  5. Click Upgrade in the pop-up window.

Mobile Backend (non-classic)


The data from the new scriptable tables do not show up in the list screen The list screen created on the remote tables data item is not showing up in the mobile (Request) app.

Mobile Backend (non-classic)


Inconsistent order of items in the list screen with multiple interleaved item streams

  1. Create a list screen with one segment and two item streams.
  2. Configure the item streams with data items, for example, incidents and wm_task, and each of them is sorted by a certain field, for example, short description.
  3. Configure each item stream to use the respective table and the interleaved field.
  4. Configure the interleaved field to use the same field which was used in the data item for the sorting.
  5. Open the list screen and observe the order of the two items which have the same value for the specified field.
  6. Use the exact same configuration, but delete the sys_sg_item_stream_m2m_segment records and create new ones (i.e. create the same mapping but with different sys_ids).
  7. Open the list screen again and observe the order of the two items mentioned above.

Expected behavior: The order of the items should be the same for a certain configuration (as long as the data have not changed). At least, changing the sys_sg_item_stream_m2m_segment records should make it possible for the user to determine a specific order.

Actual behavior: The order of the items might be reversed. For example, if the description in the two items is the same, such as 'Hello World', one being the description of an incident and the other being the description of a WOT, then sometimes the incident comes first and the WOT afterwards, and sometimes the order is reversed. No changes in configuration can be made for the order to be consistent.

Mobile Studio (non-classic)


When creating a calendar from Studio or platform UI, there is no option to create item stream time span
  1. Open Studio.
  2. Click Create applets.
  3. Choose the Calendar applet, fill in the name, and click Create.
  4. Click Submit.
  5. Click the item stream segments tab, and click New.
  6. Fill in the name and screen (name of the calendar), and click Submit.
  7. Click Insert row and click the zoom icon.
  8. Click New.

Expected behavior: There will be an option where the item stream opens or it automatically opens "item stream time p".

Actual behavior: There is no option to choose and it opens the item stream segment.

Mobile Studio (non-classic)


If Applet Launcher UI section is added and section type = Item, non-list applets are returned in the List applet dropdown

Mobile Studio (non-classic)


Stability issues occur when a user modifies the form screen's segment configuration Form screen with the embedded list screen has stability issues when users click 'modify segments' in Studio.

Mobile Studio (non-classic)


Only functions with a global context are available On Form Applet, the functions with both global and record contexts should be available.

Mobile Studio (non-classic)


Issues due to the removal of the "Active" flag

Mobile Studio (non-classic)


Issues with muli-scope enablement

Mobile Studio


Custom tables filtered from the data item table When building a new mobile applet on a custom app, the custom tables do not show up in the table field of the data item.

NLU Model Builder


The NLU model training fails if the user domain has a dot (.) in the name After creating an NLU model, the domain name is prepended to the model name. If the domain name contains a dot (.), the model training fails.



Recommended list in Initiate Conference Call Modal does not have the email id




Test Account throws an error for AWS master service accounts and patterns do not display more than 20 service accounts The refresh sub account API is failing to refresh subordinate accounts under the master account, which causes the Test Account button to throw an error.

Prerequisite: Configure an AWS Master Service Account with more than 20 associated accounts on the cloud configuration UI. Select validated account and note that only service accounts are listed.

  1. Navigate to the home page.
  2. Add Master Account credentials.
  3. Click Test Account.
It will cause the error message “An error occurred while trying to discoverand will not display any subordinate accounts.

Performance Analytics


The New Name field is not shown to users in the IE11 browser when users select a name preference other than 'Keep the source Spotlight group name' This happens when users select any of the following name preferences: 'Use a new Spotlight group name' 'Add element name to new name as prefix'/'Add domain name to new name as prefix' 'Add element name to new name as suffix'/'Add domain name to new name as suffix'.
  1. Open an existing spotlight group in the IE11 browser.
  2. Click Copy to Breakdown Element or Copy to Domain.
  3. Select the name preference as 'Use a new Spotlight group name'.
  4. Check the New Name field.

Notice that the New Name field is not shown to the user.



Issues with thread deadlocks and instance nodes restarting Applying update sets can trigger a system-wide cache flush, which can (as a result of this bug) cause a deadlock, and nodes restarting.



The table rotation in Oracle resets the sequence counter In an Oracle database, an auto-increment field works differently in New York compared with Madrid.



ElementDescriptorData objects are not deduplicated as expected The TableDescriptorCache memory usage may be higher than expected because the ElementDescriptorData deduplication is not working correctly.

Platform Runtime


OOB apps (Spokes) not shown in an instance after enabling the property

Predictive Intelligence


Multiple ML trainer records are added when multiple runtime NLU models are imported from different instances

Project Management


The date calculation is incorrect around DST change dates

  1. Create a project start date, for example, 1-Mar-2019.
  2. Create the start on the project task.
  3. Change the date to 5-Mar-2019.

Notice that the duration is incorrect as you tab out of the date field. Further changing the date also makes the end date incorrect.

Project Management


When the recalculate business rule fires, any Date/Time field (apart from start_date/end_date/work_start/work_end) set during the same transaction is incorrect

Project Portfolio Management


Demand to Project conversion is not copying the respective investment's funds and fund requests
  1. Create a Portfolio and Demand.
  2. Add some cost plans and benefit plans to demand.
  3. Create a top investment for Portfolio and add some funds.
  4. Add demand as a child investment.
  5. Allocate some funds from portfolio investment to demand investment.
  6. Navigate to demand, and click Create project.

Expected behavior: Funds received by the demand investment should get copied to project investment which gets created in background.

Actual behavior: All cost plans and benefit plans are moved from demand to project, but funds received by the demand investment is not copied to the project investment which gets created in background.

Record Watcher


Record Watcher intermittently misses record updates Record Watcher messages can intermittently fail to be enqueued if the channel is registered with a persisted responder.
  1. Install the SAFe plugin.
  2. Create PI and features and add features to PI.
  3. Create some stories and populate the feature of them.
  4. Navigate to the Big Room Planning page.
  5. Open the backlog panel.
  6. Suppose there are three stories under Feature 1 in the backlog panel, move the middle story to the board.

Expected behavior: The story is moved to the board and is no longer in the backlog.

Actual behavior: Before refreshing, the story is both shown in the backlog panel and the board.

Note: The above steps to reproduce is just one scenario where the issue is reproduced. The issue is not bound to the SAFe plugin or the SAFe Board.

Resource Management


Grid Filter issues in allocation workbench



The schedule span is not loaded in the custom schedule page

Security Access Control Lists


When the slowdown occurs, AllTermsCache can cause excessive flushing due to multiple threads loading the cache concurrently When the system is slow enough, the AllTermsCache JVM lock expires and multiple threads end up loading the cache concurrently.

Service Catalog: Service Portal Widgets


For the multi-row variable set, the date field shows today's date on a produced record with the com_glideapp_questionset_ default_question_editor macro For the multi-row variable set with a date field in it, if the record producer is submitted with a date with a format other than yyyy-MM-dd, the date variable shows today's date instead of the date selected by the user. This happens to the form that uses the com_glideapp_questionset_default _question_editor UI macro in the formatter.
  1. On the incident form, add the 'Incident Variable Editor' UI macro on the form layout.
  2. Navigate to the sc_cat_item_producer table.
  3. Open the entry for 'Create Incident'.
  4. Create a multi-row variable set on this.
  5. Create a date variable on the variable set.
  6. Change the system date preference to dd/MM/yyyy.
  7. Navigate to Service Portal and open the 'Create Incident'.
  8. Select the date as any future date or past date but not today's date.
  9. Submit the record producer.
  10. Come to the native UI and search for the incident.

Notice that the date variable is showing today's date.

Service Catalog: Service Portal Widgets


The on-load client script runs multiple times when it is defined inside the variable set mapped to the catalog item
  1. Create a catalog item.
  2. Create a variable set with the on-load client script inside it mentioning some text alert.
  3. Open this item in Service Portal and verify that the alert is displayed once.
  4. Reload the catalog item.

Expected behavior: The alert is displayed only once.

Actual behavior: The user needs to cancel the alert based on the number of times the page is refreshed.

Service Mapping


The Services Tree lazy loading stops working resulting in high memory consumption The Services Tree lazy loading stops working resulting in high memory consumption on the maps screen and maybe also on the Event Management dashboard screen.

Service Portal


Variables such as reference, choice, date/time don't work properly in the catalog Fields such as reference, choice, date/time don't work properly in the catalog opened in mobile. Users are unable to edit questions of these fields and the screen gets stuck. This behavior is observed in both iOS and Android.

Service Portal


Time is stored as 00:00:00 in record when set to a different value

Service Portal


Using a login menu item on a public homepage that links to a landing page does not redirect to the correct homepage

Service Portfolio Management


The availability report is not returning any outages when filtered by 'This year by month'
  1. Create a couple of outages on a service offering within a month.
  2. Run the data collection job for SPM.
  3. Open the service owner workspace.
  4. Navigate to the service offering.
  5. Open the report for the availability.
  6. Filter the report by 'This year by month'.
  7. Click the data points for one of the months.

Expected behavior: Any outages in that month should appear in the Outage tab.

Actual behavior: None of the outages are showing in the Outage tab.

SIAM & Vendor Management


Large result set is returning from an SQL query for the huge vendor data based on the audit, impacting the performance of the VLM application

SIAM & Vendor Management


If the Vendor Performance plugin is installed along with the VMW plugin, vendor managers cannot edit vendor fields
  1. Install the VMW plugin and the Vendor Performance plugin.
  2. Log in with the Vendor Manager role.
  3. Navigate to Vendors and click any vendor.

All the fields are read only for vendor managers once the Vendor Performance plugin is installed.

SIAM & Vendor Management


Vendor Score weight not adding up to 100%

Software Asset Management Professional


Metric attributes are not generated on the software model when entitlements are moved from one software model to another

  1. Create two software models:
    1. Model 1: Microsoft SQL Server 2009 Standard, and mention the next version as 'Microsoft SQL Server 2011 Enterprise', i.e. Model 2.
    2. Model 2: Microsoft SQL Server 2011 Enterprise Create an entitlement of type P + SA for Microsoft Per Core under Model 1.
  2. Run the scheduled job: SAM - Create downgrades/upgrades for a software entitlement. Make sure that the entitlement starts showing under Model 2.

Notice that metric attributes are not added for Model 2.

Software Asset Management Professional


Per Core metric attribute on the software model results in incorrect rights calculation

Software Asset Management Professional


The query for the spend detection labeling job is slow

Software Asset Management Professional


Modify the redirect route of samp_license_workbench_redirect to

Software Asset Management Professional


The 'Subscription Reclaimed Candidates' indicator has the filter 'Created on This year', which should be 'Closed on This year' Under the Additional conditions tab, the 'Subscription Reclaimed Candidates' indicator has the filter 'Created on This year', which should be 'Closed on This year'.

Software Asset Management Professional


Business rule in the subscription table causes an exception when a subscription record does not have a software model

Source Control Integration


Deleting and recreating a coalesced record will delete it upon processing author_elective_dir

Upgrade Engine Issues


Some upgrades fail to process base instance apps due to a permission denied error After the upgrade, not all the base instance apps are upgraded to the newest version.

Usage Analytics


Per-User subscriptions need to display a message indicating that the user should allocate users to Per-User subscriptions

Usage Analytics


Usage Analytics calls fail with status code 2 and do not persist outbound usage data on the instance since the stream is blocked Several outbound usage tracking calls fail on the instances with the following: Usage Analytics send failed due to status: 2.



De-activating and re-activating a setup topic add it to the topics list for requestors



VA designer window doesn't show the associated NLU model and the intent with it

VA-Slack Adapter


When the instance is upgraded from London to New York, Slack is not responding properly in between the topic flow

  1. Configure the Slack adapter on a London instance.
  2. Type 'hi' and choose a topic. In between the topic flow, upgrade the instance from London to New York.
  3. Continue the topic flow after the upgrade.

Notice that Slack is not responding properly. The message 'Darn – that didn't work. Feel free to give it another go.' displays.

Virtual Agent Platform


AMB clients should not resume conversations started on adapters

Virtual Agent Platform


Save the utterance in the context for both NLU and keyword utterances

Virtual Agent Platform


Unable to direct the chat to a live agent after disabling VA from the portal header

Virtual Agent Platform


Certain setup topics should not be shown to users in a choice list or the Show Me Everything menu

Virtual Agent Platform


Virtual Agent is not working after upgrading to New York

Visual Task Boards


The Label field is empty in the label_entry for label mapping

Web Services


Attaching a knowledge article to a case does not work in Agent Workspace with the custom URL Users cannot attach a knowledge article to a case (or any other record). Clicking the 'Attach' button does not cause any action.
  1. Access an instance via the custom URL.
  2. Open a case in Agent Workspace.
  3. Click Agent Assist to show related knowledge articles.
  4. Click the three dots icon and then click Attach.

Expected behavior: The article is attached to the case.

Actual behavior: Nothing happens.

Web Services


Complex object creation for top-level primitive arrays is incorrect



Remove unnecessary log messages



The time period in the stage no longer appears by default for the legacy renderer

Other New York Patch 1 information

Activate a language
I18N: Korean Translations (available)
I18N: Norwegian Translations (new)
I18N: Swedish Translations (new)
Integrating with the SAP SuccessFactor
You can also elect to synchronize employee to-dos so that employees can view and complete their to-dos from either system. For example, employees would be able to view and complete their SuccessFactors to-dos in the Employee Service Center.

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.