Thank you for your feedback.
Form temporarily unavailable. Please try again or contact to submit your comments.
  • London
  • Kingston
  • Jakarta
  • Istanbul
  • Helsinki
  • Geneva
  • Store

Istanbul Patch 5

Istanbul Patch 5

The Istanbul Patch 5 release contains fixes to these problems.

Istanbul Patch 5 was released on April 20, 2017.
Build date: 04-14-2017_1324
Build tag: glide-istanbul-09-23-2016__patch5-04-05-2017

ServiceNow QPP targets and patches are immediately available. Users do not need to request an entitlement to schedule an upgrade to these versions.

For more information about how to upgrade an instance, refer to Upgrade to Istanbul.

For more information about the release cycle, see the ServiceNow Release Cycle. For a downloadable, sortable version of Istanbul fixed problems, see KB0598935.
Note: This version is approved for FedRAMP.

Security-related Fixes

Istanbul Patch 5 includes fixes for security-related problems that affected certain applications and the ServiceNow platform. We recommend that customers upgrade to this release for the most secure and up-to-date ServiceNow features. For more details on security problems fixed in Istanbul Patch 5, refer to KB0622081.

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




Large number of rows in the Discovery Log [discovery_log] and TCP Connection [cmdb_tcp] tables can cause performance issues during an upgrade

Starting with the Helsinki release, the sys_domain and sys_domain_path fields were added to the Discovery Log [discovery_log] and TCP Connection [cmdb_tcp] tables. Upgrades of instances that have large numbers of rows in these table can lead to performance issues when the upgrade process adds and populates the new fields.

Refer to the listed Known Error KB article for details.

Embedded Help and Tours


Embedded Help REST call is degrading instance performance The Embedded Help CDN HTTP call is on a foreground client thread, which negatively impacts instance performance.

Browse to a page without any Embedded Help content.

Note the HTTP request from the client's browser takes longer then expected.

Edge Encryption



When the number of connections increases, the Edge Encryption proxy experiences extreme slowness Edge Encryption proxies can run extremely slowly. This issue occurs when concurrent user sessions drive up the number of simultaneous user connections to the instance through the proxy. There are also no logs to indicate the proxy's processing.
  1. Start up an Edge Encryption proxy and connect to an instance.
  2. Simulate users going through the proxy and using connection-intensive pages, such as Service Portal and Knowledge.

Notice that after a certain threshold, response times become extremely slow in the browser.

Transaction and Session Management



Canceling a homepage can cause CellRenderThread threads to wait indefinitely Under the right cancellation circumstances, the CellRenderThread can get stuck.

Refer to the listed Known Error KB article for details.



When an instance upgrades from Eureka/Fuji to Istanbul, the mid.version is not updated in sys_properties This leads to a false error on the MID Server form in the instance.

All Other Fixes

Problem Affected plugins Short description Description Steps to reproduce



ApprovalActionHandler is triggering an OR query (similar to PRB817692) This can be very taxing on the database.

Audit Management


GRC: Audit Management When adding profiles to an engagement, the popup's pagination is broken
  1. On an engagement, pick an engagement in the scoping phase.
  2. Navigate to the Profiles related list and click Add.

UI popup issues:

  • The arrows in the bottom right are missing; instead, there is a box that says, 'NaN rows per page'.
  • There is no way to filter or search at the top. Users cannot navigate past the first page of profiles, and they cannot add profiles to engagements.

Authentication - SSO


When a user logs out of the instance, '' appears This issue occurs when MultiSSO is configured on Helsinki and Istanbul instances.
  1. Configure an instance with the Multiple Provider SSO plugin.
  2. Using the metadata from the IdP, create a new identity provider record.
  3. Check if 'Identity Provider's SingleLogoutRequest' is defined after the import.

If it is blank, the logout request will go to ''. Note: Some IdPs do not specify a 'logout' URL.

Automated Test Framework



Automated Test Framework Screenshots are not displayed in test results due to missing extension Automated Test Framework test results are missing screenshots due to generated attachments having no extension.
  1. Navigate to Automated Test Framework > Test and create a test.
  2. Add a step to the test to 'Open a new form'.
  3. Set the form table to Incident.
  4. Run the test.

Expected behavior: The Test Result record has a JPG screenshot with a file extension.

Actual behavior: The Test Result record has a PNG screenshot with no file extension.

Cloud Management Application


Cloud Management v1: VM instances have empty names and states from AWS configuration events Incoming messages are not processed because they contain single quotes, and the CMDB is not updated. Empty names and states are also appearing. This issue can occur for users with many accounts configured with the AWS Configuration (for example, 30 accounts). Due to the heavy load of messages coming in, some events will get dropped.

Cloud Management Application


Azure discovery creates duplicate records in the Azure resource table
  1. Run discovery twice.
  2. Go to the Azure resource cmdb_ci_azure_resource table.

There are duplicate records with a Storage Blob resource type in the table. Processing these duplicate records will take a long time.

Configuration Management Database (CMDB)


CMDB Dashboard CMDB Dashboard - CMDB/PA dashboard not accessible to itil, asset role
  1. Navigate to the CMDB View Dashboard.
  2. Click the Share option.
  3. Add the asset and itil roles so that they can view the dashboard.
  4. Impersonate an ITIL user.

User receives an error message: 'The dashboard is not shared with you.'

Data Certification



Updating the related list on a data certification task does not update the related record When updating the related list on a data certification task, the related record is not updated.

Refer to the listed Known Error KB article for details.

Data Certification


Slow performance with Data Certification Users are experiencing slow performance when executing Data Certification. It can take up to 10 seconds to certify a single certification element, which negatively impacts the execution time for multiple certification elements. The data certification process can take minutes instead of seconds.

Activate the 'Data Certification [com.snc.certification_v2]' plugin.

  1. Navigate to cert_task.list.
  2. Choose a record in the Work in Progress state (with a 'percent complete' below 100%).
  3. Under Certification Task, certify elements.

The data certification process can take minutes to execute completely. This process should only take several seconds.



Domain Separation: Service Mapping and 'unified patterns' are not working correctly If a user in a domain modifies a pattern, the pattern changes its sys_ID in the domain, and the sys_ID of the global pattern is not accessible.




During Discovery, Windows 2012 storage sensor sometimes overwrites the model ID of a computer on new CIs During the discovery of a new computer, the computer OS information and the model ID are written to the CI while the storage probe is running. However, the storage probe occasionally overwrites the model ID with an incorrect value.

Discover a new Windows 2012 computer on your network. Note that occasionally the model ID shows an incorrect value on the CI.

Domain Support


'Domain - Cascade Domain - Task' business rule causes slow SQL A business rule can cause a full table scan. For large tables, this results in slow performance.

Embedded Help and Tours



Embedded Help CDN access takes a long time Embedded Help REST calls are taking excessively long when fetching help content that does not exist on the local database. When attempting to access the CDN, and if the CDN is having issues, the HTTP call can spin for minutes. This causes the user experience and front-end web browser to hang for long periods of time.

Refer to the listed Known Error KB article for details.

Event Management


Incident creation business rule race condition is affecting Alert Rule execution If an alert has a before business rule that may take a lot of time, it may cause alert rules not to work for that alert.
  1. Create a business rule that sleeps for a minute.
  2. Create an alert rule that should catch the alert.
  3. Create an alert.

The incident is not created.

Event Management


Threshold performance issues Threshold rules can take a lot of time to run. For example, it takes 22 seconds to execute a user query that occurs 3500 times a day.

Event Management


OOM on impact calculation of many services (e.g. 3000) Some users are experience recurring excessive memory usage.

Facilities Service Management


Facilities Service Management After upgrading from Fuji to Helsinki, Facilities Service Management demo records are added After upgrading from Fuji to Helsinki, demo records are added and set to Active (e.g. 12 Month Refrigerant Recharge - Carrier 48TC). Users then have to manually deactivate these records after the upgrade.

On a Fuji instance

  1. Activate the Facilities Service Management plugin. Do not activate demo data.
  2. Navigate to Service Catalog > Catalog Definitions > Maintain Items.
  3. Ensure that demo data does not exist (e.g. 12 Month Refrigerant Recharge - Carrier 48TC).
  4. Upgrade the instance to Helsinki.

On the Helsinki instance

  1. Navigate to Service Catalog > Catalog Definitions > Maintain Items.
  2. Note that demo data was added and activated.

Finance Service Management


Finance Service Management Upon upgrade, Finance templates revert to OOB settings If a user changes the assignment group in a Finance template, removes the default task, and upgrades the instance, then the assignment group reverts to the OOB value and the OOB default task returns.

Financial Management


Financial Management When a user creates a portfolio budget plan for 445 and 13 period fiscal calendars, an empty budget plan is created
  1. Install Project Portfolio Suite with Financials.
  2. Create fiscal periods for fiscal calendar type 445.
  3. Navigate to Portfolio > Create project.
  4. Create a resource plan for the project. Make sure the cost plan is created for the project and shows up in the project related list.
  5. Navigate to Portfolio Workbench > Set Target > Select Demands and Projects.
  6. Select the project by checking the planned column, and click Confirm.
  7. Click Create budget plan.

Expected result: A budget plan is created with budget items.

Actual result: An empty budget plan is created.



'sm_m2m_somodel_stmodel' table contains invalid references to non-existent so_model's and st_model's

The 'sm_m2m_somodel_stmodel' has invalid references to non-existent so_model's and st_model's, which should have been handled by SMCore's cascade delete rules on the fields. These cascade rules prevent future invalid records, but do not work on records that are already invalid.

Some ways the user can have these invalid records:
  • Come from an earlier release (before Geneva) with invalid records
  • Use update set containing invalid records
  • Scripting invalid records

Knowledge Management


When a large number of KB categories is present, opening the KB homepage is slow

In an instance with many records in the kb_category table (e.g. 27000 records):

  1. Navigate to Knowledge Base > Homepage.
  2. Click a knowledge base that contains many categories.

It takes more than 1 minute to load the page.

Knowledge Management



Knowledge Management V3 Nested Knowledge category names are truncated in the 'Filter by Type' list The 'Filter by Type' column does not display the entire nested KB category name. The KB category name gets cut off after 24 characters.
  1. In Geneva, navigate to Knowledge > Knowledge base.
  2. Open the Knowledge knowledge base.
  3. Go to the Knowledge Categories tab and click New.
  4. Create a KB category with a long name, such as "New Knowledge category Level 1" and save it.
  5. Expand the category you created and create a nested one.
  6. Create a nested KB category with a long name, such as "New Knowledge category Level 2" and save it.
  7. Navigate to Articles > Create New to create a new article with the following values:
    • Knowledge base: Knowledge
    • KB category: New Knowledge category Level 1 > New Knowledge category Level 2
  8. Save the article and publish it.
  9. Navigate to Self-Service > Knowledge.
  10. Search for the article you just published.

Note that the Filter by Type column on the right doesn't show the entire nested KB category length, which gets cut off after 24 characters.

Knowledge Management


User gets file parsing error when uploading .docx word document on Knowledge Base. Alternative text breaks the import function. If a .docx file contains pictures with alternative text, the Knowledge article import function may not work correctly.
  1. Navigate to Knowledge Base > Homepage.
  2. Click Import Articles.
  3. On the import window, drag a Word file that contains an image with alternative text.
  4. Click Import.

Expected behavior: The document is imported.

Actual behavior: The file failed to import, and the following error message appears: 'Import completed with error'.

Knowledge Management


Import issues with large Word docs: 'import completed with error' 'Unknown error', and content is lost

Large Word documents (3.5MB) are not always completely imported into the instance. A knowledge base record will be created for the imported file, but its content will be lost.

This issue occurs when using the 'Import Articles' button on the Knowledge v3 homepage.

As a user with elevated privileges:

  1. Navigate to Knowledge > Homepage.
  2. Click Import Articles.
  3. In the box labeled Drop Word files here, drag-and-drop a large Word file (3.5 MB or more).
  4. Select the Knowledge knowledge base.
  5. Click Import.

Users receive the error messages 'import completed with error' 'Unknown error'.

Legal Service Management


Legal Service Management, Marketing Service Management Fuji upgrade: Legal > Self-Service > Contract Review and Management items are reset to OOB Upon some Fuji upgrades, Legal > Self-Service > Contract Review and Management items are reset to their out-of-box settings.

MID Server


Multi-page payload support consumes a lot of memory on the MID Server When multi-page payload support was introduced, ECCSender.getOutputFilePrefix now loads the payload text field into an XML DOM. If the user has a very large XML payload, typically via imports, the XML DOM takes up a lot more memory than the original simple text field, leading the MID Server to run out of memory.




Mobile UI In mobile UI client scripts, current roles are unavailable on the g_user object

Refer to the listed Known Error KB article for details.




Android mobile app users - getDeviceType used in some UI actions for mobile returns "doctype" A function used in some UI actions returns "doctype" on the Android mobile app rather than "m." This issue prevents the UI actions that use this function from being displayed because the condition expects an "m."

Refer to the listed Known Error KB article for details.

Performance Analytics


Performance Analytics The PA Breakdowns widget breaks if the breakdown name contains special characters

Install the Performance Analytics Premium plugin.

  1. Log in as admin.
  2. Navigate to Performance Analytics > Breakdowns.
  3. Create a new breakdown with name that contains an apostrophe or ampersand (' or &).
  4. Create a widget with this new breakdown, and add it to a dashboard.

Expected result: The widget and chart should be displayed correctly.

Actual result: The widget is not displayed.

Performance Analytics


Scoresheet table can update slowly

Performance Analytics


Detailed scorecard/widget does not show chart if there are more than 1000 scores, or if the scores are far in the past
  1. Truncate all scores for 'number of open incidents'.
  2. Collect scores from 01-13-2014 to 01-02-2014 (or a time range far in the past).
  3. Collect scores for the current date.
  4. Open detailed scorecard for the indicator.

Expected result: All scores should be shown.

Actual result: Users see a message 'No data to display'.

Performance Analytics


Performance Analytics Switching back and forth in a detailed scorecard compresses the breakdown chart
  1. Open a detailed scorecard.
  2. Navigate to the Breakdown tab.
  3. Switch the chart to column.
  4. Then, select another point in time, and navigate again to the Breakdown tab.

The breakdown is compressed.

Performance Analytics Application



Responsive Dashboards, Performance Analytics Dashboard permissions migration issue after upgrading to Istanbul
If an instance is upgraded to Istanbul but the responsive dashboards functionality is not immediately activated (glide.cms.enable.responsive_grid_layout is false), dashboard permissions will be lost for:
  • Dashboards that were created after the upgrade but before activation of responsive dashboards
  • Permissions made to any dashboard after its Owner field has been populated

For new dashboards, only the dashboard owner will be able to see these dashboards after responsive dashboards are enabled. Permissions will need to be manually re-added to these dashboards.

  1. Upgrade to Istanbul but do not activate responsive dashboards.
  2. Do one of the following:
    • Populate the Owner field on the dashboard, then change its sharing permissions.
    • Create a dashboard and add sharing permissions.

      For more information, see Create a responsive dashboard.

  3. Activate responsive dashboards.

    For more information, see Enable responsive dashboards.

Note that the permission changes are not preserved, and for new dashboards, only the dashboard owner can see the dashboard.




Deleting a non-empty table that extends sys_metadata does not remove sys_metadata records for rows in that table If a table extending sys_metadata is deleted and there were rows in that table, then the sys_metadata records corresponding to the rows in that table are not removed.
  1. Create a new table that extends sys_metadata.
  2. Navigate to that table and create a new record in it.

    Note the sys_id of the the record you just created.

  3. Delete the table (without deleting the record in it).
  4. Navigate to sys_metadata and search by sys_id using the id from step 3.

Note that the record has not been removed.

Project Management


Adding or deleting project tasks in the Planning Console can corrupt the wbs and wbs_order values This occurs when the firing of business rules is disabled and when deferred save is turned off.



Report filter 'Greater than' does not work when the interactive filter value changes When using the 'greater than' condition for an integer value in the condition builder, it will lose its filter if the report is on a homepage and it is following an interactive filter.



Webkit HTML to PDF does not work when using a non-English language plugin Webkit HTML to PDF does not work when using a non-English language plugin.



Responsive Canvas Unable to select text on some widgets when using mouse device On home pages, users are able to copy data from reports by selecting it with a mouse device. However, this does not work on dashboards.
  1. Open a Dashboard.
  2. Try to select a chunk of text from a widget.

The text cannot be selected.



Responsive Canvas iFrames that do not respect the same-origin policy break the dashboard's event bus Content blocks which contain iFrames which refer to URLs which break the same-origin policy cause the dashboard's event bus to break as well.
  1. Open any Responsive Dashboard, or create a new one with several tabs.
  2. Create an iFrame content block on the dashboard and set its URL to, for instance,
  3. Reopen the dashboard and try to switch between the tab containing the iFrame content block and another tab.

Parts of the dashboard become unresponsive.

Security Incident Response


Security Incident Response Upon upgrade, customized choices in the 'State' field on the 'sn_si_request' table are overwritten Customized choices are overwritten by the file sys_choice_sn_security_request_state.xml.
  1. Enable the Security Incident Response plugin (com.snc.security_incident).
  2. Navigate to the 'sn_si_request' table and look at the State field.

    Note that there are 3 out-of-box: Draft, Assigned, and Closed.

  3. Modify several choices for the 'sn_si_request' table and State field. For example:
    1. Change the label for the OOB value Draft to Requested.
    2. For a record with the label Assigned, set inactive to true.
  4. Ensure there is a sys_update_xml record for that with the update name sys_choice_sn_si_request_state.
  5. Upgrade the instance.

The upgrade loads the sys_choice_sn_security_request_state update and overwrites the sn_si_request_state choices.

Security Incident Response


In the Security Incident table (sn_si_incident), the metric 'Create to Resolve Duration' is incorrectly configured

Service Catalog



Questions are not visible on the risk assessment survey when called within a dialog box Some users are trying to use glideOverlay as the popup on the UI macro 'Fill Out Risk Assessment'. When glideOverlay is used, the questions are being hidden.

Service Catalog



Lookup select box has slower response times in Istanbul vs. Helsinki
  1. Navigate to Service Catalog > Catalog Definitions > Maintain items and open an item.
  2. Add a new variable with a lookup select box referencing a large table (such as sys_security_acl).

    For more information, see Lookup select box.

  3. Save the variable and click Preview Item.

Response time in Helsinki: 1791ms

Response time in Istanbul: 14673ms

Service Mapping


Service Watch fix script prolongs upgrades, and security constraints prevent selection of Service field
  • Upgrade duration is significantly longer due to a Service Watch fix script.
  • During fix script execution, security constraints appear to prevent users from selecting the Service field.

The following procedure only applies after upgrades from Geneva and when the Event Management plugin (com.glideapp.itom.snac) is active before the upgrade.

In order to activate the fix for this PRB:
  1. Log in as an admin user.
  2. Navigate to System Definition > Scheduled Jobs.
  3. Search for a job named Event Management – Import impact rules.

    If you are unable to find the job, no additional action is required.

  4. If the job exists, open it.
  5. Click Execute Now.

Service Portal


Ticket Conversations widget limits journal fields to 100 entries

If there are more than 100 entries in the Ticket Conversations widget, the user cannot see the subsequent comments.

To activate the fix for this PRB in Helsinki and Istanbul, create the system property glide.service_portal.stream_readable and set it to true. When this property is set to true, Service Portal will first prune Activity fields a user cannot read before generating the journal entry stream. For details, see Ticket Conversations widget.

  1. Configure a Service Portal page with the Ticket Conversations widget.
  2. Set up an incident with an Additional Comment as the oldest journal entry, where the Caller is an end user with no role.
  3. Add 99 work notes, totaling 100 journal entries.
  4. View the incident on the Service Portal.

As an admin user, all conversations are included in the widget. As an end user with no role, only the Additional Comment is visible.

Upgrade Engine Issues


Users are noticing issues with auditing Dictionary collection entries should follow the naming convention of sys_dictionary_tablename_null.xml, but there are a few on the disk that have it as sys_dictionary_tablename_.xml. This file is named incorrectly.
  1. Make some changes to sys_user collection dictionary record (in this case, audit was enabled).
  2. Upgrade, or re-install the plugin.

Notice that audit selection was overridden by the upgrade.

User Interface (UI)


In Istanbul SLA Definitions, the Variables item is not available from the 'Show Related Fields' menu When creating start/stop/pause/reset conditions in SLA Definitions, users are unable to select 'Show Related Fields > Variables' for tables with variables.
  1. Navigate to SLA Definitions > New.
  2. For the SLA table, select a Catalog Task (or any table where variables have historically been available).
  3. For the Start Condition, select Show Related Fields > Variables.

Expected behavior: The user is able to choose variables and define an SLA against variables in a given catalog item.

Actual behavior: The variables are not displayed in the list.

Other Istanbul Patch 5 information

Service Portal: Ticket Conversations widget

Limit the number of journal entries that appears in the ticket conversation widget using the glide.service_portal.stream_entry_limit system property. The default number is 100. Change the number by navigating to Service Portal > Properties, and enter the new number in the Maximum number of stream entries displayed in Service Portal field.

Fixes included with Istanbul Patch 5

* 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.