When the user requests all data and the able is large, such request will likely not complete. Hence a paging logic is required. But as the odata service should support all tables and views, how should that be implemented? How does the subsequent read get the same data of the same transaction timestamp in the correct order etc?
The solution is to keep the resultset open for a few minutes and the skiptoken tells what resultset that is.
wernerdaehn
commented
4 years ago
When the user requests all data and the able is large, such request will likely not complete. Hence a paging logic is required. But as the odata service should support all tables and views, how should that be implemented? How does the subsequent read get the same data of the same transaction timestamp in the correct order etc?
The solution is to keep the resultset open for a few minutes and the skiptoken tells what resultset that is.