Guidance on 2004s BI options

[Former Member]

Hi all -

Need some guidance on a new (green field) implementation of NW2004s. ECC and BI in scope and EP/XI is out of scope for initial phase. Possible options include:

1)BW 3.5 back-end and front-end; BPS (with NW 2004s ECC component - not sure this is viable or reasonable).

2)2004s BI back-end; 3.5 front-end; BPS

3)2004s BI back-end and front-end; BPS

4)2004s BI back-end and font-end; IP (is this possible w/o EP?)

Currently have no BI/BW knowledge in-house. Consultants familiar with BW3.5 and inexperienced with 2004s BI. Pros and cons of above enumerated options appreciated considering: FEP/SP and upgrade issues; compatibility issues; functionality improvements 3.5 versus 2004s; stability of 2004s BI components, etc. Any other heads-up advice or options appreciated.

Thanks,

Pat

[Former Member · Accepted solution]

Pat

Suggest you wait a bit before starting with a 2004S implementation. We are in the middle of one and have found numerous errors, especially when working with the [new] core 2004 modelling objects / components. The basics just does not work yet. If you do decide to go with 2004S, then use the 3.X modelling techniques - they appear to be marginally more stable (even though we have also had our challenges with them) but will probably not be supported at some stage into the future.

Transportation system in 2004S is flaky and in some cases does not work.

From a Query / Web Application Designer point of view, 2004S is still very unstable (the new .Net f/e uses up a lot of workstation memory and falls over regularly + not all the functionality works properly).

From a BPS point of view, this is fairly good(ish) but there is nothing that you cannot do in 3.5.

I am waiting for SP09 to be applied - maybe that will correct most of the errors that I experienced (or maybe I am just an opptimist).

Strongly suggest you go for 3.5 dependent on the timelines and whether you are really going to need the "new" functionality delivered through 2004S. I know that you then have to do an upgrade and the associated regression testing, but it will certainly save you a lot of time and aggravation. As with all their roducts, SAP will get it right at some stage.

If you need to discuss in more detail, just give me a shout.

Regards

Daneel

[Former Member · Accepted solution]

Hi Daneel -

Thanks for your insight. Our project is commencing Blueprinting in October and this will run until February. Realization will take an addition 6 months therefore the actual go-live will be a minimum of 9 months away. In light of this timeline, is this risk associated with 2004s BI likely to diminish?

Also, EP will not be implemented in the initial phase. Would this impact your decision to choose 2004s over 3.5? Or, is EP simply required for some IP components?

Many thanks,

Pat

[Former Member · Accepted solution]

Hi Pat

We went live with NW2004s at the end of May 06. As we were in ramp up with SAP from December 05, we have had first hand experience of some of the issues Daneel mentioned. Although there are still a number of issues with the new tool set delivered by NW2004s, the underlying 3.5 functionality if very stable. This has allowed us to continue operating a normal BW productive service since go-live, including new project implementations as well as fixes and enhancements, using the existing 3.5 design, development and delivery methods and tools.

Given your project timelines, I really cant see any benefit in implementing a 3.5 system, as that core functionality is already available and working in NW2004s. This would then avoid the need to plan an extensive and expensive future upgrade from 3.5 to Nw2004s to take advantage of the new functionality available.

We are applying SP9 today and are hopeful that this will resolve a number of the outstanding issues mentioned elsewhere in the forum - if you are interested let me know and I will let you know how we get on.

Regarding EP, the answer really depends on what functionality you are looking to implement in NW2004s, but it is not a pre-requisite as such.

Hope this helps

Andy Bruce

[former_member194892]

Hi Pat,

I agree with Andy's reply you should go for 2004s right away. We went live end of June (ramp-up as well) and the 3.5 functionality within 2004s is very stable. Doing so will allow you to roll out the goodies on a step by step basis. Concerning EP, we went live without EP and deployed it 2 months later.

The time frame you've mentionned will give you probably 2 or 3 more SP (assuming you do a SP freeze at some point) and at that time, most of the new 2004s functionality should be pretty stable. Use some time in the blueprinting phase to give a dummy 2004s box to your 3.5 BW consultants so they can start practicing (actually this will be a good practice for your BASIS team as well!).

Hope this helps,

D.

[Former Member]

Here's a phased approach for rolling out new runtime:

[Former Member]

Pat,

The right combination really depends on the timeframe of implementation. At this point in time BI 2004s backend is fairly stable, there are few issues with 2004s frontend. If you are going to go live in near future, and taking into consideration EP not being implemented, 2004s BE + 3.5 FE would be a good choice (note you can develop using both FE's). There are a lot of good features available in 2004s for which you may endup doing some work arounds which is a cost added to the cost of upgrade (can be significant since it is major functionality change). If your time frame of implementation is > 6 months, you can defer the decision of using the FE components to a later date., since the reports are the last ones to be developed and in the mean time all the small issues with the FE components will be sorted out.

You can checkout the benefits of upgrade presentation by my colleague Dr. Berg. It will give you an idea of benefits of 2004s

http://csc-studentweb.lrc.edu/swp/Berg/articles/PM_2006_upgrade_NW2004s_Bjarne_Berg_v12.ppt

Gopal

Message was edited by: Gopalakrishna Baddela