kristofvdw_at_matt.es (K) wrote in
news:ef74fcf7.0411032327.47802a75_at_posting.google.com:
>> This is true. I view engines as separate queues.
>> You can manually add as many engines as you desire,
>> but each needs a unique name & therefore coding changes
>> upon submission to use the additional engines/names.
>
> So how do you code the distribution amongst the different queues?
> How can you know one queue is occupied by a big job, and another isn't?
>
Alternatively, create a cluster of Report Server (engines) and
submit all reports to the cluster_name. The "next" report will
be sent to the least loaded engine automagically.
Received on Thu Nov 04 2004 - 08:52:47 CST
