Hi Gurus,
We have a distributed installation with 2 dispatchers. What I am trying to figure out is what is the maximum no. of reports that can be scheduled at a time. My understanding is it is calculated as - under 'Batch Report Service'
(Number of low affinity connections for the batch report service = 2) X (Maximum number of processes for the batch report service = 10) = 20
for each server. So in total for 2 dispatchers 40 reports. So if I start 50 reports to run at a time, only 40 are in Exceute status and the rest go to Pending status. Is that correct.
Thanks
2pac
Correct - the remaining 10 reports will be queued until connections are freed up to process them.
Regards,
MF.
Thanks MF.
Also, is there a way to find out what dispatcher is running how many processes. For eg. if I schedule 10 reports to run at night, is there a way to track how which dispatcher ran which report, and how many processes each dispatcher started etc.,
Or some kind of tracing...
Or is there a way to atleast know how many shceduled reports were in Execute Status at a particular given time...we do have audit tables set up, but could not find this information, please help
This information should be available directly in the Administration Console. Just go to Launch -> IBM Cognos Administration, go to the Status tab, and select 'Past Activities' on the left. You should then see a list of reports that ran, along with the dispatcher each was assigned to.
If memory serves, the default rules for load-balancing between dispatchers are that reports will initially be routed to a single dispatcher until 50% of the connections for that dispatcher are in use, at which time further reports will be routed to the next dispatcher. This of course assumes that the weighting for each dispatcher is set the same. In your case with 2 dispatchers, having 10 processes x 2 low affinity connections each, you will probably find that 10 scheduled reports will be handled by a single dispatcher.
Regards,
MF.
wow, Thanks for the detailed information MF.