After migration from 10.1, I was getting the following error on mass-bursting a report (300+ Store Managers) by e-mail:
DPR-ERR-2072 Unable to load balance a request with absolute affinity, most likely due to a failure to connect to the remote dispatcher. See the remote dispatcher detailed logs for more information. Check the health status of the installed system by using the dispatcher diagnostics URIs. Handler trace back: [the_dispatcher] com.cognos.pogo.handlers.performance.PerformanceIndicationHandler [the_dispatcher] com.cognos.pogo.handlers.logic.ChainHandler [service_lookup] com.cognos.pogo.handlers.engine.ServiceLookupHandler [load_balancer] com.cognos.pogo.handlers.logic.ChainHandler [lb_forwarder] com.cognos.p2plb.clerver.LoadBalanceHandler
However, if I change the new burst option "Run in Parallel" to Disabled, it runs with without issues.
Note: Error was not caused by extraneous dispatchers left over from "old" Cognos version.
Running in Parallel seems like something I would want to do!
Does anyone have any thoughts about how take advantage of this?
What settings need to change?
What are the downsides?
Any thoughts/experiences would be appreicated.
Thanks in advance,
Adam.
10.2.1 runs in parallel by default where it used to run sequentially.
Because of similar errors we have set the global RSVP.BURST_DISTRIBUTION to false so by default it runs sequentially.