Servicenow-missing records issue

[sivakoteswaV]

By using ServiceNow Table API to pull data from ServiceNow There appear to be record differentials between data pulls that are not explained by new records coming in or existing ones being updated 

[Hemanth M]

Hi @sivakoteswaV ,

can you provide more details about your question?

[sivakoteswaV]

When the user retrieves data from the task_sla table using the GET method in Postman, why difference record count showing 

API Calls

Calls to the API were made by first retrieving the OAuth token and then making a query to https://ascensionesm.service-now.com/api/now/table/task_sla. Parameters used:

• sysparm_limit (40000)
• sysparm_offset (varies)
• sysparm_query
• sys_updated_onBETWEEN2026-09-01T00:00:00@2026-09-29T18:19:59^ORDERBYsys_updated_on

The JSON from the response is parsed and each row processed into an output collection; the count of processed records is added to the offset, which is being tracked alongside our normal method of progression (see below).

If the JSON response contains errors (partial records or empty rows), the valid records are processed and the total number processed is added to the offset. Errors are handled by reducing the sysparm_limit value by decrements of 1,000 at a time.

Determining Next Link

When available and when output is error-free, the link from the header value named “links” in the response is used, e.g.:

<https://ascensionesm.service-now.com/api/now/table/task_sla?sysparm_query=sys_updated_onBETWEEN2026-...>;rel="next"

[Vishal Jaswal]

Hello @sivakoteswaV 

Are you saying that when you pull data from ServiceNow using the Table API, the records returned from one extract don't match the records returned from another extract, and the differences cannot be explained by normal activity such as new records being created or existing records being updated.

Like at 9 am Today you called Incident Table API and count is 100 incidents and then at 10 am same Incident Table API being called gave you incorrect count like maybe 80.

If the above analysis is incorrect, then could you please elaborate your requirement and if the above analysis is correct, then need to learn more about how you are calling Table API and whether it used to work or a new implementation where ACL/role/sysparm_limit or sysparm_offset could be causing this mismatch.

[sivakoteswaV]

yes
iteration to iteration showing different in count