Setting Up Complex Deferred Indexing

 

In some cases, even after you set up and use deferred indexing, your QIndex workstation may still process indexes slowly. You may be able to solve this problem by creating multiple deferred queues. You create multiple deferred queues by dividing one deferred queue into multiple queues. This process is called “splitting a queue.” (For more information on this process, see Splitting a Queue.)

Setting up multiple deferred queues is a complex process that requires strong knowledge of indexing principles and SQL Advantage. If you do not have such knowledge, do not set up multiple deferred queues. Otherwise, the integrity of your indexes may become corrupt.

Do not set up multiple queues unless you have already set up deferred indexing as explained in Setting Up Deferred Indexing (one immediate queue and one deferred queue) and found the results unsatisfactory.

Use multiple deferred queues only if you meet all of these criteria:

You have substantial experience administering one immediate and one deferred queue.
Your deferred queue on the QIndex workstation is running many hours behind the indexes being processed on your technical workstation.
You have enough available workstations to be able to devote one to each new deferred queue that you set up.

Here is an example of how complex deferred indexing works:

As with only one deferred queue, your technical workstation processes the non-deferred queue. The other workstations are each dedicated to one QIndex utility processing a deferred queue.

Although queues are assigned with a number, they do not take on a priority value. Rather, the number is just the “name” assigned to the queue. Thus, queue 2 can process before queue 1, and so forth, if you have multiple queues. (Remember that queue 0 is not a deferred queue and is always indexed immediately.)

This section explains these topics:

 


© 1998-2017 Sirsi Corporation