'Standing' Database Requests

I recently posted some thoughts on how to satisfy a collection of standing queries efficiently, and drew surprising (to me) interest and responses. The problem environment is that of database updates trickling in against a large number of standing queries.

I had referred to Logicon's LMDS software, and asked for info on other solutions. But dumb me failed to forsee the level of interest, and I didn't collect these in any organized manner. So this is an invitation to re-followup for interested readers. (I'll post e-mail received on this subject.)

Most responses said, in effect, to run the series of queries using the built-in database retrieval facilities. A familiar solution. In an SQL environment, design a batch run against an index of 'new' records. But that introduces latency, and may not be acceptable in some shops.

The issue: It appears to me that the main problem is that the query environment is single query-multiple documents, while the problem at hand is that of a single document-multiple queries. Therefore, I would say, the familiar solution is badly suboptimal.

In a small database environment, no biggie. But things change in a more demanding situation, and I'd appreciate hearing about solutions/products here.

--

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
ashore_at_access.digex.net	|  Signature?  I don't need no steenkeen'
			|           signature!
			|  awww, 	Arnold Shore
			|		Annapolis, MD  (410) 849-8721