Data Sync Settings

Max Sync Waves

pulsar.sync.maxWaves

A Positive Number

24

A sync wave is generally equivalent to one month of history, and with this numeric setting you may specify how many months into the past (in terms of active records) that the algorithm should look back and fetch them. Pulsar determines active records by using the LastModifiedDate, SystemModStamp, or another datetime field (if overridden with the pulsar.sync.objects.lastModifiedDateField Pulsar Setting).

To guarantee that a user should initially sync a specific number of months of historical data, you will need to add one or two to your number depending on how you’ve configured the final wave. For example, for a minimum of two years worth of data, you can specify 26 (with the full history wave enabled) or 25 (with the full history wave disabled).

Enable Full History Wave

pulsar.sync.enableFullHistoryWave

TRUE / FALSE

TRUE

See pulsar.sync.maxWaves above for general wave information. By default, the final wave performed for an object will attempt to sync not just one month’s worth of data but all data since January 1, 1980 sorted by LastModifiedDate in descending order. This is done to optimistically download as much data as possible for the user. Since this date range is large, all or part of the request may not succeed. If it does not succeed, Pulsar does not consider this an error and continues to integrate the records it was able to retrieve.

Specifying ‘FALSE’ for this setting, will make that final wave behave similarly to the other waves and only retrieve a month’s worth of data. In this configuration, the final wave is expected to succeed or Pulsar will report a sync error.

Activity Max Sync Waves

pulsar.sync.activityMaxWaves

A Positive Number

6

This is a numeric setting that dictates how many months into the past in terms of active Event, EventRelation, Task, and Note records should the algorithm look back and fetch them.  

(Fetch all records that were modified in the past 6 months).

Object History In Months

pulsar.sync.objectHistoryInMonths

List of the pairs in the format Object API Name:Number of months (e.g:  Lead:12, Account:24, Contact:24)

24

This setting lets you control each individual object history in terms of months. 

The value should be lower than the value specified for Max Sync Waves setting. 

Objects that are subject to a stale record cleanup (due to changes in record accessibility/permissions)

pulsar.sync.recordReachability.objects

A list of object api names separated by new line, or comma. 

This setting allows you to specify which objects are eligible to scan and remove stale records from. Stale records are identified as existing records that exist on the client device but do not have a corresponding server record.  

If this setting is not specified or has an empty value, then all synced objects are eligible.

Frequency of stale record cleanup (due to changes in record accessibility/permissions)

pulsar.sync.recordReachability.intervalDays

A number greater than zero indicates that the cleanup stage should run after that many days since the last time it successfully ran. Note, to disable set to 0 or a negative number. 

0.0

This setting allows you to set the interval (as a number of days) when to perform a cleanup of potentially stale records on your device. The interval refers to the number of days since the last successful run of this cleanup stage.  This setting accepts fractional days (e.g. to specify 36 hours, you would enter 1.5). There is another frequency setting below, however, if both are specified this one takes precedence.  Note, stale records can occur in a local database if record permissions are altered after a user has performed a sync such that a user no longer has read access to a record.

Frequency of stale record cleanup (due to changes in record accessibility/permissions)

pulsar.sync.recordReachability.frequency

A number greater than zero indicates that the cleanup stage should run after that many successful syncs.  Note, to disable set to 0 or a negative number. 

0

This setting allows you to set the interval when to perform a cleanup of potentially stale records on your device. The interval refers to the number of successful syncs performed. For example, if you set it to 10, this will run every 10 successful syncs. Note, stale records can occur in a local database if record permissions are altered after a user has performed a sync such that a user no longer has read access to a record.

Force the stale record cleanup stage to run on the user's next sync

pulsar.sync.recordReachability.runOnNextSync

TRUE / FALSE

FALSE

For a description of the stale record cleanup stage, see the above setting for "pulsar.sync.recordReachability.intervalDays". If this setting is set to TRUE, all users will run this cleanup stage during their next sync (only once). The cleanup stage will not trigger again for this user unless you've specified this stage to run periodically via one of the other 'recordReachability' settings. In order to reactivate this setting for a subsequent use, make sure the value is still set to TRUE, and re-save the record to update its last modified date.

Enable Reachability on Subsequent Passes

pulsar.sync.recordReachability.enableOnOtherPasses

TRUE / FALSE

FALSE

Specifying this setting as TRUE will enable reachability to run on the second pass (in addition to the first pass) only if local changes were pushed during the first pass.

Initiating a sync means that Pulsar will have at least one pass through all the sync stages. However, another pass will be performed if, during the first pass, updates were either downloaded or local changes pushed to SFDC. Since the reachability stage can be a time consuming stage, it will normally only run during the first pass.

Also, the reachability stage runs before the stage where Pulsar pushes local changes to SFDC. Therefore, if those changes might render some records unreachable, they will remain on the local device until the next time reachability runs.

Object Filter Limit

pulsar.sync.initialFilterLimit

ObjectAPIName:Number

Default:50000

Object_API_Name:Number

This is a setting that dictates how many records Pulsar should fetch for each table (e.g: Lead, Account, Contact etc.). By default, the app caps at 50,000 records for each of the objects it syncs and the admins can increase the limits if they see fit.

Auto Sync Setting

pulsar.sync.enableAutoSync

TRUE / FALSE

TRUE

This particular setting dictates how the app would run the sync process. If the value is set to true, the app will automatically perform a bi-directional sync without requiring the user to initiate the sync. If this is set to be false, the user will have to manually choose to Sync in order reflect the changes on the device and the server appropriately. The manual sync can be initiated by tapping on the blue ‘Sync Now’ button on the Settings page as well as by clicking on the circle that appears on the bottom toolbar, which then will display a popup window with the blue ‘Sync Now’ button.

Allow/Deny Sync through 3G/4G

pulsar.sync.enable3GAutoSync

TRUE / FALSE

TRUE

This option would permit or not permit auto sync to run when the device is connected to the data network. Some companies might want to restrict the data usage and in those cases, the admins can use this setting to control the sync behavior.

Force a Layout Sync

pulsar.sync.enableOneTimeSessionRefresh

TRUE / FALSE

FALSE

In order to save API calls, the catch-up sync does not sync the layout changes every time. Layout sync only runs if the user clicked the 'Refresh Settings' button from the Settings page. Or, if you would like to force a layout sync, enable this setting after making layout changes to the org.  When Pulsar next performs a sync on users devices, this will including syncing the layout changes. This forced metadata sync only happens once for each user after this setting is created/updated. 

Auto Sync Time Interval

pulsar.sync.autoSyncIntervalMinutes

Number of minutes

120

Some customers asked us to allow a sync interval time as opposed to running the sync every time the app comes to foreground. For example with the below setting, the sync algorithm will check to see if there was a successful sync in the last X minutes and if there was one, it will not initiate the sync process. If X minutes indeed have passed since the last successful sync, it will initiate sync.

Sync Data Filter using a "Where" Clause

pulsar.<sobject type>.syncFilter

A SOQL format where clause (without the 'Where' keyword).

This is a per-object sync filter to further control which records get synced. Note, it is technically possible to create/edit a record that would not get subsequently picked up due to the object's sync filter.  We would pick up the server delete, but not any server update.

Since the UserInfo SOQL syntax for obtaining the current user's id is not available outside of APEX, you can instead use this Pulsar defined constant:  @@CurrentUserId

Example: 

K: pulsar.Account.syncFilter

V: Account.Name = 'Apple'

Allow Base 64 Download

pulsar.sync.base64download.enabledObjects

Comma or New Line separate list of object API names  (e.g: Attachment, StaticResource)

Pulsar by default doesn't fetch the Body field (the actual data that constitutes the file) of Attachment, StaticResource, and other similar objects during the sync process. This setting enables that explicitly. Customers using this should know that there could be quite a lot of data and advise users to use a WIFI connections for the initial sync.

NOTE that without this setting specified, Pulsar will still download the data on an individual basis when attempting to view attachments (while online).

Allow Sync and display of Rich Text Field Images 

pulsar.sync.images.richtextimagefields

Comma or New Line separate list of objectAPI:FieldAPI Names  (e.g: Lead:UploadImageTest__c)

Pulsar by default doesn't display images supported by the Rich Text Field data type. This is a feature that could come in handy when you are implementing a field survey or maintenance log. User can upload an image right into the field that is displayed as part of the record. 

Sync Objects - Last Modified Date Field Override

pulsar.sync.objects.lastModifiedDateField

A list of objects separated by newlines where each line has the object api name and the name of the DateTime field to use instead of LastModifiedDate in this format:  <object API name>:<field API name> 

By default, the LastModifiedDate field is selected as the DateTime field that is used to determine when a record has been modified.  This is a built-in Salesforce field that is updated automatically for your organization, and is used by Pulsar to download new and updated records.  Although not recommended, you may override this field per object with your custom DateTime field if necessary.  One possible reason you may want to do this is because your organization runs batch jobs that frequently update many records and result in the LastModifiedDate field getting updated.  In this scenario, you may decide that you don't want all the records updated in this fashion to be considered 'modified' and consequently re-downloaded by Pulsar in the next sync.  Perhaps you only want to refresh only those records where a subset of fields are updated.  One way you might do this is by creating a workflow rule that updates a custom DateTime field when specific fields are updated.  If you implement this setting, please keep in mind that you are responsible for maintaining when and how your custom DateTime field gets updated.

Account:My_Self_Managed_LastModifiedDate__c
Contact:My_Self_Managed_LastModifiedDate__c 

Pulsar Sync Object List

pulsar.sync.objects

A list of objects separated by commas or newlines

Add the names of the objects you want to sync to this list and then copy/paste that into the PulsarSettings Page.

Important Note: If a relationship directive setting is also found, this setting will be ignored as the other one takes precedence.

The list of objects to sync can be separated by comma, semicolon or newline. Also note that, you need to include the standard objects along with the additional objects. In other words, this is the absolute master list Pulsar honors and syncs the objects you specified here

Account
Case
Contact
CurrencyType
Event
Lead
Opportunity
OpportunityContactRole
OpportunityCompetitor
OpportunityStage
RecordType
Task
TaskStatus
User

[Add list of additional objects like Product and Custom objects at the end of the above list] 

for eg: 

TestCustomObject__c (add the api name of the object)

Pulsar Sync Objects - Relationship Directives

pulsar.sync.relationship.directives

Look at the Notes section for a sample setting

None

Specify this setting to indicate that you would like Pulsar to sync your data using object relationships. You define relationship 'directives' where each directive tells Pulsar what object to sync and how to sync it. Each directive is defined as having a depth number starting at 0 and increments by 1. This depth represents both the order that the directive will be synced (0 is first) but also the dependency involved. For example directives defined at depth 2 depend on the objects synced at depth 1 which in turn depend on the objects synced at depth 0. Depth 0 has no dependencies. Additionally, directives at depths greater than 0 are required to define the dependency from the prior depth (object and field name). 

There are two forms for specifying directives (see below) for depth 0 objects and for other depths.

Delimiters used in parsing this setting:

1st level delimiter | (pipe) separates distinct query directives
2nd level delimiter ; (semicolon) separates a directives components
3rd level delimiter : (colon) separates components into subcomponents

Format for starting directives (depth 0):

0 ; <object type that we're syncing> ; <optional SOQL where clause> ; <optional sync window usage>

Format for all other directives (defines the dependency at the previous depth):

<depth greater than 0 > ; <target object type that we're syncing> : <optional target field name (default Id)> ; <source object type from previous depth> : <optional source field name (default Id)> ; <optional target SOQL where clause>

Limits and caveats:

• This setting takes precedence over the pulsar.sync.objects setting if both are specified.

• You may specify up to 32 depths of directives, however, we've rarely seen more than a few depths needed at most.

• Currently, any time this setting changes (regardless of the significance of the change), Pulsar will perform a longer sync similar to an initial sync to make sure any data on the user's device is valid under the new relationship hierarchy defined by this setting.

Content File Directives:

• To sync content file objects (Content Version, Content Document, Content Document Link), we recommend you follow the best practices in our Enable Salesforce Files page.

• While you can explicitly specify Content object directives, we don't recommend it at this time, with the exception of including them at depth 0.

0 ; ServiceAppointment ;
Id IN (Select ServiceAppointmentId from AssignedResource WHERE ServiceResource.RelatedRecordId = '@@CurrentUser.Id') AND
SchedStartTime <= NEXT_N_DAYS:45 AND
SchedEndTime >= LAST_N_DAYS:45 AND
Status NOT IN ('New', 'Scheduled','Canceled') ; NoSyncWindow
|
0 ; Account ; Primary_Address_Country__c='@@CurrentUser.Country'
|
0 ; Lead ; Lead.OwnerId = '@@CurrentUserId'
|
0 ; Case ; OwnerId='@@CurrentUserId'
|
0 ; AppExtension
|
0 ; FieldServiceMobileSettings
|
0 ; MobileSettingsAssignment
|
0 ; Organization
|
0 ; Product2 ; CID__c != '' and CID__c != null
|
0 ; WorkOrderStatus
|
0 ; WorkOrderLineItemStatus
|
0 ; ServiceResource ; RelatedRecordId = '@@CurrentUser.Id'
|
1 ; AssignedResource:ServiceAppointmentId ; ServiceAppointment
|
1 ; WorkOrder ; ServiceAppointment:ParentRecordId
|
1 ; WorkOrderLineItem ; ServiceAppointment:ParentRecordId
|
1 ; Account ; ServiceAppointment:ParentRecordId
|
1 ; Account ; ServiceAppointment:AccountId
|
1 ; ServiceTerritoryMember:ServiceResourceId ; ServiceResource
|
2 ; Asset ; WorkOrder:AssetId
|
2 ; Asset ; WorkOrderLineItem:AssetId
|
2 ; ServiceResource ; AssignedResource:ServiceResourceId
|
2 ; Location ; WorkOrder:LocationId
|
2 ; Location ; WorkOrderLineItem:LocationId
|
2 ; WorkOrderLineItem:WorkOrderId ; WorkOrder
|
2 ; Account ; WorkOrder:AccountId
|
2 ; ServiceTerritory ; ServiceTerritoryMember:ServiceTerritoryId
|
3 ; Location:RootLocationId ; Location:RootLocationId
|
3 ; Product2 ; Asset:Product2Id
|
3 ; ServiceTerritoryLocation:ServiceTerritoryId ; ServiceTerritory
|
4 ; Location ; ServiceTerritoryLocation:LocationId
|
5 ; ProductItem:LocationId ; Location

Sync In Progress Banner

pulsar.sync.enableSyncInProgressBanner

TRUE / FALSE

TRUE

When enabled, a red banner appears over all screens when a sync is in progress advising the user not to leave the app until sync has completed.  

The banner is shown by default because sync can only run while the app is the foreground app.

Enable Field Service Sync Setting

pulsar.sync.enableFieldService

TRUE / FALSE

FALSE

When enabled, Pulsar will sync all Service Report Templates (ServiceReportLayout) available to the user. These templates are available to the Pulsar Platform developer through the Pulsar javascript API.

Forced Work Offline Mode (with Sync Enabled) Setting

pulsar.sync.enableForceOfflinePermanently

TRUE / FALSE

FALSE

When enabled, Pulsar will remain in "Forced Offline" mode, with the following exceptions:

• manual and automatic sync will still be available.

• Users will not have the option to control the offline/online status or the autosync status from within the app. These must be controlled via settings.

Disable syncing listviews for certain object types

pulsar.sync.listViews.disabledObjects

Comma or New Line separate list of object API names  (e.g: Account, Contact)

This setting allows Pulsar to skip syncing the listviews for the specified object types. When an organization has a large number of listviews for a particular object type (e.g., 1000+), this can significantly slow down sync. Without this setting, all available listviews for object types in the sync objects setting are synced.

Remind users to sync periodically

pulsar.sync.notification.remindMeInXMinutes

Number of minutes

1440

This setting allows Pulsar to use device notification feature to remind users to sync Pulsar to make sure their local data stays in sync with the server.

You can specify 0 in the value field to turn off sync reminder notifications. 

Enable Content Document Direct Sync

pulsar.sync.contentDocument.enableDirectSync

TRUE / FALSE

FALSE

If syncing the content document object, specifying TRUE for this setting will query/sync the ContentDocument (CD) object directly as part of the sync object list.  Otherwise, only CD records that are associated with a previously synced ContentDocumentLink object will be downloaded.

Enable Chronological Changes

pulsar.sync.enableChronoChanges

TRUE / FALSE

FALSE

When disabled, all local changes are sent to the server in 'standard order' which is Pulsar's historic way of recording and transmitting changes. This means that when syncing, Pulsar will push all local deletes first, followed by all local creates, followed by all local updates. Note, that when multiple updates to a particular record are performed, the changes are all captured together into a single record and then sent to the server as a single update request.

When enabled, all offline creates, updates, and deletes are recorded and sent to the server chronologically (i.e., in the order they were input into Pulsar). When multiple updates to a particular record are performed, each update is captured and sent separately. 

The benefit of enabling this setting is that Pulsar can mimic exactly how a user would have entered the data and properly handle certain business dependency/sequence logic that may exist on the server.  However, this may result in additional sync time when sending the local changes to the server as Pulsar potentially now has to send multiple updates instead of one update per record. 

Configure Conflict Resolve Policy (Global)

pulsar.sync.conflictResolvePolicy

USER_DECIDES / SERVER_WINS / CLIENT_WINS

USER_DECIDES

A sync conflict occurs when the user makes a change to a field in a record while offline and that same record's field is updated to a different value on the server since the user's last successful sync.  When the user next performs a sync, they are notified that a sync conflict exists and is asked to choose between the local device's value or the server's value before completing sync.  Once the user selects the value, that is the value that will ultimately be set for both local and server values for this field.

By default, the user will be prompted for every record's field that is in conflict.  This setting can be specified to override this default behavior (USER_DECIDES) and automatically select either the local device's value (CLIENT_WINS) or the server's value (SERVER_WINS) for all fields in conflict.

If your organization requires more fine-tuned control, please use the object/field specific version of this setting.  Please note that when both settings are used, this global setting applies as the default and is overridden by the specific object/field setting.

Configure Conflict Resolve Policy (Object/Field Specific)

pulsar.sync.conflictResolvePolicy.<object type>

A list where each line has a field name followed by a colon, followed by the conflict resolve policy string.  For example:

Field_1__c : CLIENT_WINS
Field_2__c : SERVER_WINS
Field_3__c : USER_DECIDES

Please see the global conflict resolve policy setting for a general description.

Configure Duplicates Policy

pulsar.sync.create.allowDuplicates

pulsar.sync.update.allowDuplicates

Comma or New Line separate list of object API names  (e.g: Account, Contact)

These will allow the duplicate to be created/saved even if alerts are enabled via duplicate rules on the Salesforce server, without displaying any alerts or interrupting sync on the Pulsar side. The setting is a blanket policy for each object in the list. If you have set up duplicate rules on Salesforce, and you have not set up this setting, Salesforce will return errors on save or sync when it encounters a duplicate via configured matching rules.

Please Note: This setting will not merge any records. 

Enable Composite API

pulsar.sync.enableCompositeAPI

TRUE / FALSE

FALSE

Please see Composite Graph API below for more details as they are very similar. This API has a significantly lower maximum operations per request than the graph API, but it is recommended if there are any implicit dependencies between objects in your org. This is because only this API can guarantee that the order of execution matches the order of the requests sent.  Typically, implicit dependencies are created via server side triggers.

Composite API differences from Composite Graph API:

• Maximum of 25 operations per request

• Guarantees sequential processing

If both Composite and Composite Graph API are enabled at the same time, Composite API takes precedence.

Enable Composite Graph API

pulsar.sync.enableCompositeGraphAPI

TRUE / FALSE

FALSE

When enabled, Pulsar uses the Composite Graph API to send as many local changes as possible in one request to the server. When this setting is disabled, Pulsar sends creates, updates, and deletes one change at a time per request.  The benefits of composite graph include fewer API calls and faster times to send changes.  

However, not all changes are eligible for a graph request. Here are some of the known SFDC limitations per graph request:

• A maximum of 500 operations per request

• Some objects require headers that are not available with graph requests, so these must use the SOAP API and be sent separately

  • Case, Event, EventRelation, Task

• The maximum number of different object types is 15

• The maximum number of updates per record is 12

In addition to the previous limitations, there are also a few Pulsar limitations at this time due to the special handling required around certain objects. For example, Pulsar will mark the following as graph ineligible: 

• File object types: ContentDocument, ContentVersion, ContentDocumentLink

• Chatter object types: FeedItem, FeedComment

• Any object create containing a base64 field

• Any object create/update containing a pending base64 upload within a rich text field

• If using Pulsar Setting, duplicateCheckField, graphs will be constructed to avoid mixing object creates that supported duplicate checking with those that do not. The first object create encountered by a graph is what determines what that graph will allow. Subsequent graphs will then alternate as necessary.

When one of these limitations is found within your local changes, we send the graph request with the eligible changes found up to that point.  After that, we send that ineligible change separately. The process is then repeated until all your changes have been sent.

SFDC treats the changes in the graph itself as an all-or-none type of request.  This means that if one change fails, then the entire set of changes in that graph are rolled back.  Pulsar will subsequently prompt the user to correct the issue and, when corrected, a subsequent sync resends a newly calculated graph of changes.  This process repeats until all changes are successfully sent.

*Composite Graph guarantees order of execution in the group of pushed changes only for explicit dependencies. If there is no explicit dependency, then it's possible that SFDC might execute some those requests out of order to improve performance.  If your org has created implicit dependencies via triggers, please use the Composite API instead.

Maximum Updates Per Object

pulsar.sync.maxUpdatesMap

ObjectAPIName:Number

With this setting, organizations using relationship directives can specify a maximum number of updates per object during a catch-up sync. If an object is found to have exceeded this number, then the object is marked for re-sync. What this means is that the normal catch-up queries are not run and instead only relationship queries are run for this object type using the initial sync window criteria.  This can be beneficial to organizations that sometimes have an extraordinary amount of changes to a particular object that do not necessarily apply to the user's relationship directive hierarchy. Consider using this setting to improve sync performance if your organization is seeing catch-up syncs that are exceeding initial sync times.

This setting is currently only intended to be used with relationship sync enabled orgs and applies only to objects with directives at depths greater than 0.  Also, with this setting specified, Pulsar issues a count() query on SFDC only for those specific objects to determine how many updates will be downloaded during the current sync run.

Connect Timeout Seconds

pulsar.sync.connectTimeoutSeconds

A positive number

60

The time in seconds Pulsar will wait for connection acknowledgement from Salesforce servers.  This is not the total time to send back all query data, just the time it takes to connect/begin getting the data.

Generally, on 3G or better networks that are not overloaded you can expect an initial response to a Salesforce query in under 10 seconds.

Most organizations will not need to change this setting.

For organizations that do exceptionally complex queries, the Salesforce servers themselves may take a long (> 1 minute) timeframe to respond.

If you see the Pulsar data sync process repeatedly fail with timeout errors for multiple users across different networks (home/work/field), consider increasing this number.

Prevent Duplicate Create and Update Transmissions

pulsar.sync.duplicateCheckField

The field api name you want to use to have Pulsar check before attempting to re-send a create/update that may have already succeeded. The field identified here should be a text field supporting at least 50 characters and be marked as an 'External Id' to have Salesforce automatically index this field.

After Pulsar succeeds in sending an offline created or updated record to the server, Pulsar typically gets a success response.  Occasionally, however, the success confirmation is not sent.  This is typically due to a server timeout, but can also be caused by any interruption at this stage. When resuming a sync where that interruption happened, Pulsar does not know whether that last create/update was successful and assumes it was not to avoid loss of data. The consequence of this is that a create may be retransmitted causing a duplicate to be created or an update may be sent again potentially causing undesirable effects.

To mitigate this, this setting automatically generates a unique transmission Id per unique record and saves this Id into the field specified by this setting (must also exist on the object) and sends the record to Salesforce. If there is an interruption at this point, the subsequent sync will first query Salesforce for that temporary Id. If a record is found, Pulsar will not re-send the create or update and will continue as if the record change was just successfully sent.

To enable duplicate checking and handling, you must:

• Specify this setting with a valid field api name

• For every object that you want to guard against duplicate create or update operations:

  • Use an existing field or add a new text field with the same name as the one specified by this setting to the object 

  • Make sure that the text field can support at least 50 characters and mark it as an External Id

Bypass Failed Changes

pulsar.sync.enableBypassFailedChanges

TRUE / FALSE

FALSE

When pushing your local changes during a sync, Pulsar will interrupt the sync immediately if a change was not accepted by Salesforce to allow you to address that failure before continuing. Enabling this setting allows Pulsar to skip any failures to continue pushing local changes. Note that when a change is skipped, Pulsar will also skip any subsequent changes that are directly related to that skipped change. A change is directly related if it has a parent/lookup or child relationship to the skipped change. Note that when a related change is skipped, that also means any subsequent changes with a direct relationship to that change is now also skipped.

For example, say you created an account, a contact off of that account, and a case off of that contact. Also, say that the account create failed. In this scenario, all three changes would be skipped because they form an unbroken chain of relationships to that failed Account create.

After attempting to send all changes, the user will be left with only failed changes or changes related to that failure and be prompted to address the first error encountered.

This setting does not apply if your organization is using the Composite API or Composite Graph API Pulsar settings. That is, the behavior will be the same as when this setting is false or not specified.