Missing records with REST Get calls when using sysparm_offset in the queryIssue Try to fetch incidents using REST API (/table/{table_name} endpoint) with sysparm_offset and sysparm_query (to fetch events of specific time) parameters but some records are being skipped. For Example,1. At time T1, If we have fetched 8 records, i.e. [1, 2, 3, 4, 5, 6, 7, 8] all having time as T1, then we'll store Time=T1, Offset=8.2. At time T2, On SNOW (server) side, Record no. 3 gets updated, and 9 and 10 are newly added, hence now the sequence would like [1, 2, 4, 5, 6, 7, 8, 3, 9, 10] (Notice here that, Snow isn't keeping the older record in place)3. At time T3, We request to fetch new/updated records using the last stored time (To avoid duplication). So, The request would be time>=T1 and offset=8.Now, In this request, the expectation is it will give the records after the 8th record. So, We'll get records [9, 10].Since Record no. 3 is also updated and it is now not the same as previously fetched ideally it should also be fetched but it is being skipped.CauseThe behaviour of the platform is expected as per the design and cannot be changed.When a record gets updated in short time spans ( REST call at T1 and by the time we pull the next set at T3 , one of the earlier record pulled at T1 gets updated at T2) by design we will keep moving in out of the offsets and limits with every update, which will be dynamically keep moving forward in time filter criteria + offset limit criteria that your job uses. The above record's dynamic nature will in-turn impact the other records too even though these other ones have not constantly getting updated. The impact would be again in a way that these un-updated records keep pushing backward on the offset which in turn will lead to getting skipped from your job's respective query criteria and in result set too. For example, assume there is a record INC18515355 that has last updated time as 10-16-2020 13:31:08. When your job query to pull records between 13:30:00 to 16:30:00 at 15:31:09(offset 0 – limit 1000), the incident record INC18515374 is at offset 1000 and INC18515355 is at offset 1001. Whereas when your job query to pull records between 13:30:00 to 16:35:00 at 15:33:00 (offset 1001-2000), the incident record INC18515374 moved to offset 3001 as it got the time stamp updated as 15:32:34 and INC18515355 gets pushed backward to offset at 1000. So your job that ran at 15:33:00 for offset 1001-2000, might ,miss the INC18515355 as its current offset is at 1000. ResolutionOverall for the active records which are open and where you are trying to pull the data, the behaviour which is stated in this KB is expected from the platform as per the design. This is expected when there are quick inserts or updates happening on the records when the complete REST query is executed. Workaround is to pull large offsets of data to minimize the issue or use some kind of logic at the third party (which is trying to pull the data from servicenow) based on the response from servicenow to ignore duplicates ( that are updated at T2) and make sure you get all the records which are created at T2 as well when you query at T3.