A live search request is similar to the simple search request. The
difference is that a live search request does not complete. Instead, the
search returns all the matching records to the client, and then keeps the
HTTP request open. As new data is ingested into the repository, matching
results are immediately sent to the client.
Important
Live searches can be terminated at any time, and do not work with
aggregate functions and operators. For more information on the operation
of Live Queries, see
Live Search Operation. This may mean
that a live query fails, or 'times out' during execution and results may
fail to be returned without a clear indication of the reason. Using the
Query Job API may lead to a more reliable and reproducible result set.
For more information, see Running Query Jobs.
To execute a live search request, add the
isLive
parameter to the request object with the value
true. For example:
The results match the filter will be returned instantly, but the HTTP
connection will not be closed. Any additional results will be returned as
they are ingested: