AnsweredAssumed Answered

query performance, RepositoryServiceSoapPort.fetchMore()

Question asked by multi_io on Oct 1, 2007
Latest reply on Oct 26, 2007 by nuga
What's the best way to deal with queries that may return thousands of result rows? Or should those be avoided altogether? I have a simple cm:folder here that contains some 15,000 children, and a RepositoryServiceSoapPort.queryChildren(theFolder) causes the client to time out and close its end of the socket after a minute; after 5:30 min or so, the server apparently tries to write back a result and, obviously, fails.

I see that there is a "fetchMore(String querySession)" method in RepositoryServiceSoapPort, but I can't figure out how to use it to fetch the results in more than one network roundtrip – QueryResult has a querySession field, but when (and if) the QueryResult is returned it already contains all the result rows, no?

I wouldn't mind if large queries take longer to complete, but it should at least be possible to do complete them somehow. Do you have to use the query(store,query,includeMetaData) for more advanced use cases like this?

Thanks in advance.

Outcomes