We have a mixture of Deski and Webi reports scheduled to run daily. They used to all run with few problems but recently we have been having huge delays on our deski reports (100’s still waiting to run from the previous night), with all Webi reports running as planned.
I’ve been spending a lot of time migrating our key Deski reports to Webi over the last few weeks to make sure they run, but we have so many more deski reports migrating them all just isn’t possible.
Is anyone aware of issues with the Desktop Intellgence servers that could be causing these delays or experienced anything similar?
Do the Webi servers take priority and could they be sucking up resources from the Deski job server?
WebI and DeskI jobs run on separate job servers, so it’s not likely a matter of WebI reports taking priority.
There are many, many things that could cause this. I would start by monitoring the Instance Manager as the jobs are processing. Are only a couple of DeskI jobs running concurrently? Maybe the maximum number of jobs needs to be increased, or maybe one DeskI job server is having problems.
We’ve always had the Deski job server set to 5 concurrent jobs which used to handle the workload fine. We now have less reports but these huge delays; reports which would run in a coulple of mins are now hitting 30+ mins
I’d need to speak with IT but I don’t think we have more than 1 physical server, I think its all set up virtually. Would it then be possible that the webi job server is somehow taking precedence?
I guess it’s possible, but only at the OS-process level. You could certainly test it by shutting down the WebI job server and see if the DeskI jobs run faster. But I’d be pretty surprised if they did.
If the queue is full (that is, five jobs running consistently) then I would check to see if the queries are remaining active in the database.
Not really sure how to check if the queries are still active in the database other than just looking at the active jobs in CMC (or using one of the excellent repository documentors I downloaded from here a few years ago)? I’ve never done anything with trace logs if that’s what your suggesting?
After migrating a few more of our longest running reports to webi yesterday and restarting the server before end of play everything seems to have run OK this morning… No doubt it’ll start playing up again at some point though!
For now I’m just happy not to be taking calls from confused managers
(BOB member since 2002-08-29)
(BOB member since 2002-08-29)