I’m guessing they are a unique code or counting sequence for every record when it’s created, and if its modified, the sysrevid is changed.
I’m trying to figure out if I can use those fields to guide incremental loading of a datawarehouse so as to speed things up, i can import records with a sysrowid that’s > than the last one refreshed since records don’t all have a date/timestamp.
SysRowID is a unique value for that record (in fact it is likely unique for any record) and never changes. SysRevID shows what revision the record is on. Not sure it’s logic, but it can be used to compare if a change occurred.
SysRevID
Is this
The purpose of SysRowID is to provide one Identity column for any record. Think about Parts,TranGLC their natural primary key has multiple parts having just one makes it easy to reason about in a generic way like extended UD tables and handling replication type scenarios.
SysRevID is like a timestamp, changing on each update to a record - using these together allows you to model optimistic concurrency and easy interfacing to other systems.
So your question about using these to do incremental loading of a datawarehouse - yes there are many ways but a simple one is like this
