COGNOS 10 Upgrade from 8.3

Currently in the process of performing a COGNOS BI upgrade from 8.3 to 10.1.

Instead of performing a big bang approach upgrade on the existing server, have deployed a new server, and performed a clean install of 10.1. Setup the new 10 environment – we now have two COGNOS environment 8.3 & 10. Once setup, imported the entire 8.3 package (which I had previously exported from 8.3) into 10. Walla – all my reports are packages appear as there were in 10. One major task to remember – it has brought along all schedules – the first step post import, is to use the new COGNOS 10 Schedule suspend feature and suspend all schedules for the day and maybe for the next few days whilst all testing is performed. Alternatively you could disable all schedules

Have found a few of the dimensional reports are failing with strange error messages – shall post the error messages 2morrow.

NO MORE CUBE SWAP….. with 10 (and 8.4 I think) transformer now takes care ofh swapping cubes, by using a MDC file pointer that points to the current MDC file, and even better is the ability for it to delete old version – you can specify the number of versions to retain – finally a very much needed enhancement.
Transformer login with multiple namespaces works  – always had issues in 8.3, where autologin to namespaces would fail, seems to be working fine with no issues at all on 10.
Transformer speed improvement!!! it now uses the available memory to build cubes in RAM rather than disk I/O. The reduction in cube build times is very noticable. In a couple of cases, my cubes builds have gone from 26 minutes to 4 minutes!!!

I still reckon SQL Analysis Services build faster cubes!!

Unknown's avatar

About Dharmesh

Have always loved travelling and will continue to pursue this love. Favourite Travel quotes: “Traveling – it leaves you speechless, then turns you into a storyteller.” – Ibn Battuta “The world is a book, and those who do not travel read only one page.” – Saint Augustine
This entry was posted in Uncategorized and tagged , , , . Bookmark the permalink.

Leave a comment