Cube Status Page

Question

I need to create a cube status page and need ideas. I thought I would post this and see maybe if someone has an idea.
Details:
Everynight, we start running batch and all our Essbase users are logged out of the cubes. We run batch through out the night. Our cubes become available the following day around 8:00 AM. I need to create a .NET/website that will show the status of the batch as cubes become unavailable/available. For example, I would like to list the names of our cubes, their status(avail/not avail), timestamp of when it became available, and most importantly ETA(Expected Time Arrival). I would like to create this page on a SharePoint site. We have about 25 Essbase cubes and are on version 9.3x. We have both ASO and BSO cubes. The servers we use are UNIX.
In conclusion, my goal is to create a site for cube availability status page and display the above fields(cube name, status, enable time, ETA, etc.,). Does anyone have an idea on how to accomplish this especially the ETA?
Thanks,

Answer 1

Hi,
ETA Expected Time Arrival, for this you need to calculate how much of time the transaction would take ( when i say transaction , it means Load or calc process).If they are new cubes, which have been run so far , then you need to simulate calculation to estimate calculation time(refer chapter 55 :'optimizing calculations').If you have been running cubes ,then you know already by now how much time the cubes have been taking.
Just pay attention to the following things:
1. How many cubes have been run simultaneously
ex: If yoru XYZ cube when run around 2 PM is taking 30 minutes for its processing.We cannot conclude that it would take 30 minuts irrespective of anything. The XYZ cube ,when being processes , there miht be other cubes which are simultaneously being processed ( which would impact the processor and RAM sharing..and other resources).
2. Take care also of the specific cubes run as per your batch to conclude the ETA , when different cubes are run simulataneouly ,again ,basis the cubes nature it might take time .
3. For any specific month/quarter ,for any particuar cube , the data to be loaded is more ( might be because of some business reason) , do factor in that point . It might happen one in a quarter of year.
For now , i see these points on the top of my head .
Out of curiousity, is it a requirement for your operational comfort or any other reason.
Sandeep Reddy Enti
HCC
http://hyperionconsultancy.com/